Slashdot Mirror


It's Official - AMD Buys ATI

FrankNFurter writes "It's been a rumour for several weeks, but now it's confirmed: AMD buys ATI. What implications is this merger going to have for the hardware market?" In addition to AMD's release, there's plenty of coverage out there.

508 comments

  1. Tomorrow by glebd · · Score: 3, Funny

    Intel buys Nvidia. Let the war continue!

    1. Re:Tomorrow by d3bruts1d · · Score: 3, Insightful

      Good lord... I hope not.

      *shudder*

    2. Re:Tomorrow by C0vardeAn0nim0 · · Score: 4, Interesting

      IIRC there's a mutual deal between ADM and Intel (part of the setlement of a lawsuit) that allows one company to use the other's technologies. that's what allowed AMD to integrate SSE 1/2/3 in athlons and Intel to integrate AMD64 in pentium4/xeon.

      if they both buys graphic chipsets companies, does this means nvidia's technology on ATI GPUs and the other way around ?

      or will they shield the newly aquired techs from the setlment ?

      --
      What ? Me, worry ?
    3. Re:Tomorrow by Slashdot+Junky · · Score: 1

      Although I don't know for sure, I suspect that the IP sharing was limited to specific technology covered in court. If the sharing was across-the-board, it would be hard for either party to gain a competitive edge over the other.

      Later,
      -Slashdot Junky

      --
      .
      Landfill Mining Co.
      Managing the (Un)natural Resources of Tomorrow
    4. Re:Tomorrow by ichigo+2.0 · · Score: 4, Informative

      http://contracts.corporate.findlaw.com/agreements/ amd/intel.license.2001.01.01.html

      As far as I can tell this deal only covers patents made before 2001 (section 2). I could be wrong though, not very good at legalspeak, and didn't read the entire contract. AFAIK they have another cross-licensing agreement as well, but it only covers all x86 extensions and improvements. This is the deal that you're probably talking about as SSE and AMD64 are x86 extensions. So to answer your question: no they would not need to share tech acquired from ATI.

    5. Re:Tomorrow by vhogemann · · Score: 3, Insightful

      Consider for a moment, that Intel does provide usable OpenSource drivers for their Video Chipsets.

      --
      ---- You know how some doctors have the Messiah complex - they need to save the world? You've got the "Rubik's" complex
    6. Re:Tomorrow by LordOfTheNoobs · · Score: 2, Interesting

      Hard as in has to have the better implementation of the various technologies rather than just exclusive rights to them?

      --
      They're there affecting their effect.
    7. Re:Tomorrow by Tim+C · · Score: 1

      But then the agreement would most likely allow the other party to copy that implementation - any advantage gained would be purely temporary.

      True, that would spur further innovation, but then so does proprietary tech that the other company can't have at all - they still have to innovate to compete.

    8. Re:Tomorrow by kestasjk · · Score: 1, Interesting

      So the second best CPU maker and the second best GPU maker pair up? Unless AMD+ATI have an ace up their sleeve this won't be good for competition.

      --
      // MD_Update(&m,buf,j);
    9. Re:Tomorrow by utopianfiat · · Score: 1

      Patents up to 2001 and until the agreement expires.
      Cheers.

      --
      +5, Truth
    10. Re:Tomorrow by ivan256 · · Score: 2, Insightful

      Ok, considered.... And dismissed.

      I hate how people write off ATI and Nvidia as Open Source scrooges since their drivers are closed. The reality is that their code isn't all home grown and they couldn't open source it even if they wanted to. The copyright and patent holders on their licensed technologies wouldn't let them.

    11. Re:Tomorrow by utopianfiat · · Score: 1

      It's Pfficial!

      --
      +5, Truth
    12. Re:Tomorrow by evil_Tak · · Score: 2

      Actually, the reason NVidia has always given for not releasing their driver source is that it would reveal too much about the architecture and design of their cards.

    13. Re:Tomorrow by hobo+sapiens · · Score: 1

      It is Pfficial. Evidently /. editors have developed a strange aversion to using the preview button!

      --
      Use the Preview Button!

      --
      blah blah blah
    14. Re:Tomorrow by Andy+Dodd · · Score: 1

      That's because their chipsets are so limited that they don't run into any issues with patent licensing that require their drivers to be closed source. Even an old GeForce 4MX outperforms the GMA series integrated graphics chipsets in performance and functionality. For all practical purposes, the GMA series chipsets are 2D-only. (Yes, they officially support 3D features, but the 3D functionality is so limited that it may as well not be there.)

      If I recall correctly, this is why ATI went from being fully documented to binary-only for full functionality under Linux - The first issue was S3 Texture Compression, which is why for a while UT2003 (or was it 2004?) only ran on NVidia cards - Until the binary ATI drivers for Linux were released, the open source ones didn't (and legally couldn't) support S3TC.

      --
      retrorocket.o not found, launch anyway?
    15. Re:Tomorrow by kimvette · · Score: 4, Insightful

      Actually, their competitors are unaffected because:

      1. They have large enough staff to decompile and perform clean reverse engineering of NVidia's drivers, e.g., one team analyzes the decompiled code and takes notes (without copying code of course), another team designs improvements and implements based on that analysis

      2. Their competitors own electron microscopes, making analysis of the chip internals relatively simple.

      Now tell me: why are the likes of NVidia and ATI keeping their products undocumented and their drivers closed?

      And to counter your argument: what happens in two years when ATI and NVidia decide your card is too old to support, and yet it still performs very well but you NEED the features in the latest kernel and latest x.org? Go ahead, buy a new video card -- oops, nope, sorry, they changed slot specs again, and PCI Express cards are no longer available because PCI-X finally gained market share in the consumer market and PCI-E ended up as short-lived as VLB did in the VLB vs. PCI war.

      (do I expect PCI-E to die? No, it was a hypothetical example showing the potential problem with proprietary drivers)

      --
      The Christian Right is Neither (Christian nor right). See: Matthew 23, Matthew 25, Ezekiel 16:48-50
    16. Re:Tomorrow by mallan · · Score: 2, Informative

      And to counter your argument: what happens in two years when ATI and NVidia decide your card is too old to support, and yet it still performs very well but you NEED the features in the latest kernel and latest x.org?

      Um, NVIDIA still supports the TNT on Linux, and that card was released in 1998.

      Just because proprietary software vendors can be evil/irresponsible/negligent/whatever doesn't necessarily mean they will be. I think that, by now, NVIDIA has proven that they're a responsible player in the Linux arena.

      --
      "Good people drink good beer"
    17. Re:Tomorrow by fossa · · Score: 1
      The reality is that their code isn't all home grown and they couldn't open source it even if they wanted to.

      Yet the fact remains that the drivers are not "free", and regardless of the excuses I'll happily take my dollars elsewhere.

    18. Re:Tomorrow by Com2Kid · · Score: 1

      Um, NVIDIA still supports the TNT on Linux, and that card was released in 1998.


      They don't support the GeForce 256 though.

      They removed support from their drivers for older GeForce chips. Maybe they added them back in and I didn't notice. :-p
    19. Re:Tomorrow by Anonymous Coward · · Score: 0

      This deal really makes me sad. It's one of the few Canadian companies that is an international success, so of course it's bought out. First Canada's (and North America's) oldest store, The Hudson's Bay Company, is bought by an American and now this.

    20. Re:Tomorrow by mallan · · Score: 2, Informative

      Use the "Legacy GPU" drivers. GeForce 256 is still supported.
      http://www.nvidia.com/object/unix.html

      --
      "Good people drink good beer"
    21. Re:Tomorrow by rmdyer · · Score: 1

      "Now tell me: why are the likes of NVidia and ATI keeping their products undocumented and their drivers closed?"

      From someone who works in the "real" world...money. It's all about money. The items you mentioned, about decompiling the source and electron microsopes, well that takes time and money, and lots of it. It's the same reason that you don't do it yourself. While some specific patented technologies can be looked into and "reverse engineered" if it is felt that the outcome of doing so will be more viable, it just doesn't make any sense to do it on a large scale, it will just cost too much.

      Very few /.'ers seem to understand the concepts of time or money. The understanding of those concepts seem to come with age. Guess what? We don't live forever, and we don't have all the time in the world to sit down and reverse engineer a system.

    22. Re:Tomorrow by Mister+Whirly · · Score: 1

      Why would it even matter?? If they gave away all their design secrets, who is going to compete with them?? When was the last time another video card producer besides ATI and NVidia has any mareket share at all?? Seen a Matrox or S3 card lately??? Even if another company started now, good luck.

      --
      "But this one goes to 11!"
    23. Re:Tomorrow by agallagh42 · · Score: 1

      Sorry, but you are in this case full of crap. :p

      EM64T is a full implementation of the AMD64 instruction set, and is fully capable of running 64bit windows and linux OSes.

      --
      Carpe Cerevisi - Seize the Beer
    24. Re:Tomorrow by TheRaven64 · · Score: 2, Interesting
      You are confusing EM64T and PAE. PAE extends the address space to 36-bit, allowing 64GB of RAM to be used (although individual processes can only access more than 4GB by jumping through a lot of hoops). This has been around since the Pentium Pro.

      EM64T is Intel's version of x86-64[1]. It is slightly different from AMD's implementation, but most code compiled for one will work on the other.


      Can we please stop calling it AMD64? It's a small number of extensions to IA32 - smaller than SSE - and AMD introduced it as x86-64.

      --
      I am TheRaven on Soylent News
    25. Re:Tomorrow by arodland · · Score: 1

      I hate how people write off ATI and Nvidia as Open Source scrooges since their drivers are closed. The reality is that their code isn't all home grown and they couldn't open source it even if they wanted to.

      I think the grandparent's point was at least Intel can make drivers that work and don't suck donkey dick unlike ATI. And they even open-source them into the bargain. They were one of the first to release open 802.11a/b/g drivers for Linux, they contribute on open 3D drivers, etc. NVidia's drivers aren't open, but at least they're properly maintained. ATI, on the other hand, releases drivers that are consistently about 2 years behind the curve on Xorg compatibility, they break cards without warning, they use silly nonstandard build procedures, and their drivers still don't work as often as not. Yes, there's been some improvement -- but the ATI drivers today are suitable to the state of Linux in, say, 1998, not 2006. It ain't good. :)

    26. Re:Tomorrow by Cutie+Pi · · Score: 4, Informative

      You've obviously never tried to reverse engineer a chip using an electron microscope. If so, you wouldn't be saying it is "relatively simple". You obviously don't realize that modern chips have upwards of 8 layers of wiring stacked on top each other, that cover the local interconnects and transister wiring (poly and active area). Now, let's say you can decompose the chip, layer by layer, snapping photos as you go. With today's 65 and 95nm processes, you can only see a few transistors at a time, given the resolution and field of view of the microscope. These chips have millions of transistors on them. How long do you think it would take to cover a sizeable area of the chip, to the point that you could gain some useful information about it? At best, you could probably look at a few latch circuits at a time... That's like looking at the grains of sand on a beach and trying to map out the coastline of Hawaii. By the time these alleged reverse engineers are done, Nvidia would have released two new architectures. Money and time are much better spent creating rather than copying.

    27. Re:Tomorrow by kimvette · · Score: 2, Interesting

      If they don't have the staff on hand to do this sort of thing, I'd be very surprised. Besides, my point isn't that they should be doing this, it's that their true competitors can afford to do this (and probably DO decompile drivers and analyze the GPUs very closely) so why slight third-party developers who are actively willing to work to develop drivers for your products for FREE?

      It's amazing how so many slashdotters who work in the "real" world are purposely argumentive for solely the purpose of being difficult.

      --
      The Christian Right is Neither (Christian nor right). See: Matthew 23, Matthew 25, Ezekiel 16:48-50
    28. Re:Tomorrow by squiggleslash · · Score: 4, Insightful

      That's complete bollocks. Nothing related to copyrights and patents prevents ATI and nVidia from releasing the specs to the X.org/XFree developers. Nobody's ever said "ATI must release the code to fglrx", not least because we know it's so awful we would want to start from scratch. ATI and nVidia's poor reputation in the Free Software community has to do with their refusal to provide non-NDA-encumbered documentation.

      --
      You are not alone. This is not normal. None of this is normal.
    29. Re:Tomorrow by martinultima · · Score: 1

      My machine's right now using a 16MB Vanta AGP 4x card, and it works just fine – I will admit, I have no clue when it was released, but it works just fine (use their legacy drivers). This is with kernel 2.6.16.19, X.org 6.9.0 – (relatively) brand-new software environment. Will admit the hardware's a bit old – Duron 700MHz, 256MB RAM – but either way it works just fine. And I have around three or four systems here with Intel integrated graphics that work just fine, too.

      Meanwhile, I've got another machine with an Athlon 64 and ATI Xpress 200 integrated graphics – works decently enough with the Radeon driver, but no 3D acceleration whatsoever... environment is 64-bit build of kernel 2.6.17.4, X.org 6.9.0, all the required 32-bit packages, shared memory enabled... tried at least ten times to get ATI's drivers to install, and it worked just fine, except that it didn't – X wouldn't show anything, the kernel module built but wouldn't load... the eerie thing is, I was actually working on that just yesterday, too. Kind of odd.

      Totally off-topic by now, but just thought I should mention anyway...

      --
      Creative misinterpretation is your friend.
    30. Re:Tomorrow by Anonymous Coward · · Score: 0

      Sorry about the Hudson Bay company.

      Though with ATI, I think it shows how well Canadians compete in difficult high-tech markets. Vid chips are as bad a microprocessors for requiring and enormous amount of high-powered engineering just to stay in the game. ATI has been in the game for quite a while. When I was doing video streaming work, the capture/recompres software they included with their $400 vidcard was every bit as good as systems I'd been quoted $2500 for. If not for their previously bad support of Linux I wouldn't have any other kind of card in my system.

      That said, I hope they keep the best engineers in ATI when the do the acquisition. I would love to see what could be done with a "dual-core" sized die, where one half was an FX-62 and the other was an 1900 XTX.

      Cheers.

    31. Re:Tomorrow by Anonymous Coward · · Score: 0

      releases drivers that are consistently about 2 years behind the curve on Xorg compatibility, they break cards without warning, they use silly nonstandard build procedures

      I think a large part of that is because they have a hard time keeping good engineers. From what I hear from recruiters (being a device driver engineer keeps you in contact with the recruiters that know this stuff), ATI has fairly high turnover in that department.

    32. Re:Tomorrow by vhogemann · · Score: 1
      And to counter your argument: what happens in two years when ATI and NVidia decide your card is too old to support, and yet it still performs very well but you NEED the features in the latest kernel and latest x.org?


      Well, it just happened. The last two iterations of the FGLRX driver broke my ATI Radeon 9200, and looking at the bugreports it looks like the entire R200 family is affected. When prompted for a solution, ATI said that R200 is no longer supported.
      --
      ---- You know how some doctors have the Messiah complex - they need to save the world? You've got the "Rubik's" complex
    33. Re:Tomorrow by Anonymous Coward · · Score: 0

      Intel last week pulled certain ATI contracts/rights. Worse anybody at ATI with knowledge that's tainted will be a future problem for AMD.

      From the sound of it AMD expects to lose large parts of ATI's business over the next year. Don't ask me to explain this deal. It sure doesn't make sense to pay the price AMD is paying for just a stump of ATI but that's what they are doing.

    34. Re:Tomorrow by Anonymous Coward · · Score: 0

      One thing though. I as perceive it, and I don't know much on the subject, true, the interface protocol needed to communicate with the chip doesn't need to have anything to do with the chip's architecture, or the reasons why its performance is so, and so. Or very little anyways. Well documented protocols would only mean that more platforms could support their hardware. They are not in the business of selling software, as I understand, but only hardware. The more easily it is to support it the more people can use it. They may even get "free" help in writing and fixing drivers which is not what they want to spend most of their money on, obviously. In that case, the monetary incentive would also speak in favour of open protocols, whouldn't they? The performace perks does not need to be tied into the protocol.

    35. Re:Tomorrow by runderwo · · Score: 1

      Sorry, that's bullshit. NDAs are not, and never have been, the problem. The problem is hardware that is impossible to get documentation for at ALL, such as any R300 or above chip from ATI, or any NVIDIA hardware. The vast majority of open source graphics drivers were developed under NDA!

    36. Re:Tomorrow by SyncNine · · Score: 1

      At least I was up-front about the possibility of being completely full of crap! :)

      Thanks for the correction, information and for the wiki link. For the record, it's been that kind of Monday all around for me. I should just shut up and go home and skip Mondays from now on.

      --
      To the darkened skies once more, and ever onward.
    37. Re:Tomorrow by squiggleslash · · Score: 1
      The vast majority of open source graphics drivers were developed under NDA!
      By definition, they clearly weren't. Try again.
      --
      You are not alone. This is not normal. None of this is normal.
    38. Re:Tomorrow by Anonymous Coward · · Score: 0
      The vast majority of open source graphics drivers were developed under NDA!

      By definition, they clearly weren't. Try again.

      Actually there are a couple of GPLed drivers that were developed using documentation under NDA. I know that doesn't make much sense since anyone can read the source code afterwards but there are apparently some suits in some companies that are quite happy with the arrangement of releasing GPLed source code that was made using their NDAed documentation for whatever reason.
    39. Re:Tomorrow by aichpvee · · Score: 1

      AMD buys NVidia. Let the war be over!

      --
      The Farewell Tour II
    40. Re:Tomorrow by Anonymous Coward · · Score: 0

      Don't forget Tim Hortons. Man, we own you guys! Next up: MegaBlok.

    41. Re:Tomorrow by Trelane · · Score: 1
      I hate how people write off ATI and Nvidia as Open Source scrooges since their drivers are closed.
      Scrooge or not, their implementations (at least on Linux) fail the Just Works test. It's why I chose Intel video (despite its poorer performance) for a desktop I just built, and it's the reason I hope to buy an Intel video laptop in the next refresh. Particularly on notebooks, since I've had a heck of a time convincing fglrx to let me suspend to ram/disk. And I know I'm not alone on this problem, and I have heard that Intel suspend to disk/ram Just Works. Given my experience with this last PC, I don't doubt it at all.
      --

      --
      Given enough personal experience, all stereotypes are shallow.
    42. Re:Tomorrow by Fulcrum+of+Evil · · Score: 1

      How about we get a binary device interface that doesn't change much? I know it opens up a rat's nest of troubles, but for 3d chips, it's probably the best way. Otherwise, how would you ever support nvidia except on specific kernel versions?

      --
      "We returned the General to El Salvador, or maybe Guatemala, it's difficult to tell from 10,000 feet"
    43. Re:Tomorrow by chadruva · · Score: 1

      I find silly that companies use patents as excuses to no release documentation of their hardware to do drivers, I understand that some "IP" may be licensed from 3th party, but saying that releasing these documents will help the other companies is SILLY, got patents? use them, they protect your inventions soo other vendedors can't get hold of your new stuff without your permision.

      They may even have an extra revenue stream by licensing their card designs or techniques.

      --
      C-x C-c
    44. Re:Tomorrow by Anonymous Coward · · Score: 0

      Typical Slashbot logic. Reverse-engineering is possible in principle, therefore it's trivial, therefore it's a waste of time not just handing over all your IP to your competitors.

      It's a very long way from you owning an electron microscope to it being "relatively simple" to analyze the internals of a 100M-transistor chip ... and there is absolutely no way that you can consider this process more "relatively simple" than if nVidia just published their design documents for free online. You're talking about thousands of man-hours versus a couple of minutes to download a PDF. Security through obscurity may be technically flawed but it works a whole lot better in practice than security through open publication of all your secrets.

      Also, forgive my ignorance of Linux, but why would a kernel update break a driver? Don't they have, like, interfaces and stuff that are locked down? I'm running an XP kernel from 2005 but for some reason my old Win2K drivers still work. If Linux developers think they have so much vendor support that they can afford to change the driver interface willy nilly maybe someone should take a big stick and talk to them about that? It might explain the poor driver support up to this time.

    45. Re:Tomorrow by ivan256 · · Score: 1

      Users who need the performance or cutting edge features are, unfortunatly, unable to happily take their dollars elsewhere. That allows them to get away with not caring about the one in 100,000 who do what you are doing.

    46. Re:Tomorrow by runderwo · · Score: 1

      Obviously, you've not a clue what you're talking about. The NDA applies to the documentation, not the developed driver. The current S3 and ATI drivers only exist because of NDA programs.

  2. What implications is this merger going to have by Anonymous Coward · · Score: 0
    1. Re:What implications is this merger going to have by Anonymous Coward · · Score: 0

      Not much, except you'll now have to sign an NDA before reading it.

  3. Linux by Anonymous Coward · · Score: 0

    Maybe we'll get some decent linux drivers now...

    1. Re:Linux by PunkOfLinux · · Score: 5, Interesting

      I was thinking the exact same thing. But I think that, if done correctly, this could really solidify both AMD and ATI as market leaders. If AMD pressures their new acquisition to create a half-decent set of Linux drivers, then this will all be worth the hassle. Or even if they convince ATI to open up the specs, that would be ok too.

    2. Re:Linux by mikeisme77 · · Score: 1

      What would also be nice is if they also pressured ATI to create good drivers in general. Their Windows drivers suck as well, but I concur that it's more important to have driver's that aren't a royal pain in the ass (it's the number one thing that makes me think twice before installing another new Linux distro or makes me wait to upgrade to the latest version of my current Linux distros--although Linspire handles my ATI card fairly gracefully, but I have other issues with Linspire...) to get to work in Linux than to have non-crappy drivers for Windows. Still, it would be nice if we could have both... or maybe they should just open source the drivers and that way both problems can be fixed at little to no cost to them?

    3. Re:Linux by dfjghsk · · Score: 4, Insightful

      right.. because as we all know.. Linux support == market leader.

      They may become a market leader for Linux desktops (GPU's aren't needed in servers where Linux is popular).. but Linux desktops are only 1-2 percent of the desktop market...

      so even if they gain all of it.. they still won't be a market leader in GPUs.

      --
      Help me take back Slashdot. When did 'News for Nerds' become 'FUD and Conspiracy Theories for Extremist Nutjobs'?
    4. Re:Linux by slocan · · Score: 1
      [...] but Linux desktops are only 1-2 percent of the desktop market...

      Today.

      But a year from now, who knows? Two years from now? Foresight from AMD/ATI could win it market share in a growing (and maybe promising) market?

      What would be the costs (money, IP etc) of developing a free driver, against leveraging the untapped demand for the added value of a GPU with a free driver?

    5. Re:Linux by dfjghsk · · Score: 1

      when linux gains a notable market share, then AMD/ATI can/will develop their drivers.. The cards (and the drivers) that will run on Linux when it is popular haven't even been produced yet. Why develop drivers for an unpopular desktop OS for cards that will be obsolete by the time it really matters.

      Besides.. I've been hearing people like you for many many years.. every year is the year of Linux on the desktop. and when it isn't this year, it's predicted to happen next year. Do you expect a company to spend their money based on that alone?

      --
      Help me take back Slashdot. When did 'News for Nerds' become 'FUD and Conspiracy Theories for Extremist Nutjobs'?
    6. Re:Linux by dnoyeb · · Score: 1

      IIRC, historically ATI has been the market leader in graphics card penetration. They have had the laptop and OEM market locked up for a long time. Its only in aftermarket sales that NVIDIA has been strongly competitive. And recently NVIDIA has made inroads into the laptop market. Traditionally, video out has been the market for ATI.

      But I was not aware that NVIDIA was the market leader.

    7. Re:Linux by robw810 · · Score: 1
      If AMD pressures their new acquisition to create a half-decent set of Linux drivers, then this will all be worth the hassle. Or even if they convince ATI to open up the specs, that would be ok too.

      Um, didn't you get that backwards? The *best* thing would be to get the specs opened up; in fact, I would argue that anything less is unacceptable, but that's just me...
      RW
    8. Re:Linux by Anonymous Coward · · Score: 0

      The thing is, if your card is clearly superior for Linux use, you'll sell to more than just Linux desktop users. There's also the Linux dabblers, who will prefer ATI over NVidia "just in case" they want to try switching to Linux in the future. Then there's the friends and families of the Linux geeks, who rely on the geeks for buying advice. The Linux users are more likely to recommend the hardware they use themselves.

      I'm not suggesting that securing the Linux desktop market would be a huge coup, but it would have a bigger impact than just that 1-2%.

    9. Re:Linux by Zardus · · Score: 1

      Um, you do know that nVidia is one of the companies with the very best trackrecord for Linux support? All their cards are fully supported, dating back to the TNT, and the Linux drivers perform as well as the Windows drivers do.

      This issue alone has made me exlusively an nVidia customer (#1 feature of a laptop for me: nVidia GPU), and anytime a friend of mine that has even a remote chance of switching to Linux needs a new video card, I highly recommend nVidia over ATI.

      --
      You can mod your friends, you can mod your nose, but you can't mod your friend's nose.
    10. Re:Linux by slocan · · Score: 1

      I am not saying that this year, or the next, or the other is the year of Linux on the desktop. But it could be. Who knows?


      There were companies who thought that the Internet was hype. And missed the plane. Having to scramble to catch up then, and still having a tough time with innovative competition -- by still clanging, some would argue and I would agree, to old ways of doing business and developing software.


      The same goes when new business models arise. Some may be the future, some aren't and some can no longer walk. Who knows? On the other hand, there were (are) the bubbles. The real hype. I agree with you about that.


      But my point was that there can be a market for GPUs with free drivers. They have an added value. One that makes them different from all other offers. (A (bad?) analogy would be music without restrictions, having more value than DRMed music). And the value isn't only for those who use Linux, but for those with demanding or specific graphics needs (in all OSes) unattend by the closed source drivers.


      The point is that maybe the same rationale that goes behind open standards, open formats, free software etc may make the same sense for hardware, regarding stability and having control of the hardware you buy. There might as well be a demand for that. And new and innovative solutions, models, new uses to old and current products etc might arise. Who knows?


      It might take some time to build solid free drivers and a surrounding community. But being drivers one of the main causes of instability, many can benefit. Or even in ways we cannot forsee. And maybee it wouldn't take son long.


      It may be a small market now, the demand might not be that huge. But it may just be untapped. Maybe with a working product, the demand will arise. Build it and they will come :-)


      My point also is that at some stages companies can decide to invest on a new product or practice, that could make them pioneers, with the possible benefits of being the only one to offer a given product or service. (There are risks, of course.)


      I don't know all the issues of such a decision to AMD/ATI, but given what I know, I think there can be commercial virtue in openness, which can also build trust (that has commercial value too).


      Of course I might well be wrong. But who can't? ;-)

    11. Re:Linux by Cocoshimmy · · Score: 1

      Here are some facts the slashdot community vehemently refuses to admit:

      -ATI and NVidia have a a pretty equal marketshare as it stands for PC Graphics. It sways back and forth but relatively equal

      -In terms of performance, it sways back and forth. Until just a few WEEKS ago, ATI was the performance king with the X1900XTX running in crossfire. So whenever nvidialinux fanboys come out and say nvidia hardware is better, it pisses me off.

      -In terms of drivers, I mean the drivers used by 99% of gamers (Windows XP), they both make very stable and decent drivers. WAAAAY back in the day, ATI had bad drivers, but a lot has changed since then. They have a unified driver archicture now(Catalyst) just like nvidia does and it performs very well. ATI evolved a long time ago, but the slashdot community hasn't and still thinks that ATI drivers stink. Oh and btw, because I know someone will bring it up, I don't care if nvidia has better linux drivers and I don't think most real gamers do either. So in terms of what matters most to consumers and to the company, bad linux drivers don't count against ATI's driver record.

      Don't believe me? Look it up yourself. Check out tomshardware, arstechnica, or whatever. If you're lazy, check wikipedia.

      So ENOUGH with the linux fanboy shit. They don't support linux for good reason, because it doesnt make good economic sense. How many games run natively on linux anyways? So exactly how many sales will that drive? How much will it cost to develop those drivers? It will certainly cost more to develop those drivers than the benefit they'll get from sales.

    12. Re:Linux by squiggleslash · · Score: 2, Insightful
      So ENOUGH with the linux fanboy shit. They don't support linux for good reason, because it doesnt make good economic sense. How many games run natively on linux anyways? So exactly how many sales will that drive? How much will it cost to develop those drivers? It will certainly cost more to develop those drivers than the benefit they'll get from sales.

      Except you've gotten it so screwed up it's not funny.

      What the Free Software community has been asking for is for ATI and nVidia to quit developing their own proprietary drivers, and release the documentation necessary for us to develop our own. If they want to release the source to their own, then that's great, but we're not impressed by them.

      So actually, ATI and NV are doing the exact thing you're suggesting is too expensive for them, and not going the cheap route. They are developing drivers. The Free Software community ("linux fanboy shit") is asking the exact opposite: we're asking for the information we need to develop the drivers ourselves. We don't want them to develop proprietary garbage (or rather, we don't care if they do), we want our own drivers.

      --
      You are not alone. This is not normal. None of this is normal.
    13. Re:Linux by Lord+Kano · · Score: 1

      Gaining or losing the Linux market isn't going to make or break any company, but there a a lot of Linux based 3D workstations out there. Spending a hundred thousand dollars on OSS driver development might bring in a couple of extra million dollars worth of sales.

      LK

      --
      "Hi. This is my friend, Jack Shit, and you don't know him." - Lord Kano
    14. Re:Linux by Anonymous Coward · · Score: 0

      Just try installing ATI's drivers on any distribution other than Red Hat or SuSE – especially one of the smaller ones – and you'll see why nVidia is so much more popular.

      By the way, last time I checked, there were quite a few major Linux games – Doom, Quake, even Unreal Tournament all have Linux versions available, and there's more everyday.

      Stop with your Windows fanboy shit.

    15. Re:Linux by level_headed_midwest · · Score: 1

      I totally agree with you (Gentoo 2006.0 on an Nforce 4 with a 6200TC PCIe GPU.) I certainly do recommend hardware that is also Linux compatible to my friends and acquaintances as even though they will be running Windows on it; giving companies that support Linux a bone is a good thing. Also, parts that have drivers in other OSes tend to be of better quality and more stable in Windows as there can't be a reliance on Windows code in the driver- i.e. it can't be a cheap win-peripheral or it won't work in any other OS worth a damn. Drivers for real hardware-controlled devices tend to be more stable and robust as there is far more independence on the host software to make the peripheral work.

      --
      Just "gittin-r-done," day after day.
    16. Re:Linux by floydman · · Score: 1

      I can say your half right...
      you would never imagine the market available for visualization clusters, involving parallel rendering, and visualization distribution, remote visualization, etc...
      Not only that, but in a few months (maybe a year or so), i can see major gaming companies targetin the new gaming box from AMD...and thats gonna be only the beginning

      --
      The lunatic is in my head
    17. Re:Linux by Cocoshimmy · · Score: 1

      Wow. 3 Games for Linux!! Thats incredible!! Seriously, thats nothing compared to whats available for windows which is in the thousands. There is not even a comparison. All the games you mention are available for windows PLUS 1000+ others. There may be more linux games everyday but there are WAY more windows ones everyday. Why? Most developers use directX or something similar. It's not windows fanboy shit, its reality. Nvidia may be popular amongst linux users because of their support, but in terms of the total market, it makes little difference.

    18. Re:Linux by deinol · · Score: 1

      and not going the cheap route. They are developing drivers.

      If I recall, at least for Nvidia, the binary driver blob is almost exactly the same for linux as for windows. It's only the wrapper they have to re-write for linux. So while I agree that they should just go ahead and give out specs, I don't imagine it is a very expensive process to maintain the linux blobs they have. Especially now that they already have the pipeline and ground work for the linux wrapper.

      --
      Got Apathy?
    19. Re:Linux by Cocoshimmy · · Score: 1

      Ok, first i'd like to point out that I have nothing against open source drivers or linux advocates in general. The "linux fanboys" I am referring to are those who choose believe that JUST because ATI does not develop good linux drivers they suddenly are doomed to have poor marketshare and bad hardware. Clearly this is not how things have panned out.

      Anyways, back to the topic at hand. In principle it would work nicely for the open source developers if the hardware manufacturer did release their hardware specs publically.

      However, there are some problems with that:
      -It costs money (engineering hours) to assemble all the necessary documentation needed to develop the drivers
      -A large part of said documentation would have to be stripped anyways because it would likely contain company confidential information and/or documentation licensed from a third party. This results in higher legal and engineering costs.
      -Releasing such documentation helps the competition by exposing their hardware architecture. This can obviously be very costly to the company.
      -Additional costs come from them having to update their documentation regularily as changes are made.
      -They'd be getting almost no financial benefit from the extra costs incurred. Open source drivers would certainly not be as fast and stable as the proprietary ones. This I understand is debatable, but remember that ATI/Nvidia have large teams of well trained engineers working full-time to develop and optimize their drivers. Also, they can easily collaborate with their hardware teams (assuming they are not part of them) and have access to ALL the proprietary hardware documentation and hardware resources (graphics cards, oscilliscopes, and other lab equipment) they need.

      With these extra issues and costs associated with opening up their hardware specifications, it's no surprise that neither company does it.

  4. Don't really know.. by Roy+van+Rijn · · Score: 4, Interesting

    ..if this is a good thing or not. It might be good for the development and cooperation. Better integration == better graphics/faster machines?

    But on the other hand, this could split the market and get things like todays uncompatible browsers. (Which is VERY annoying somethimes)

    And we have a psychic

    1. Re:Don't really know.. by babbling · · Score: 3, Interesting

      There's a slight chance that AMD might be smart and release hardware specs for ATI cards, or make the drivers Free Software. If either of those things happen, this would be a very good thing, in my opinion.

    2. Re:Don't really know.. by Anonymous Coward · · Score: 0

      Me fail english? That's unpossible!

    3. Re:Don't really know.. by Roy+van+Rijn · · Score: 4, Insightful

      Well, that could be a very good thing. Those specifications will also help driver-makers a lot. It might also help to get the linux drivers which are pretty poor for ATi at the moment.

      The AMD-fans/nerds are more linux-minded then Intel (IMHO), and AMD probably knows this. They can really make a business-blow by releasing this, in the mind of open-source.

    4. Re:Don't really know.. by Homology · · Score: 1

      > There's a slight chance that AMD might be smart and release hardware specs > for ATI cards, or make the drivers Free Software. If either of those things > happen, this would be a very good thing, in my opinion. AMD has, so far, been very open source friendly by releasing hardware documentation. ATI used to be more open source friendly than they are now. Hopefully the merger will lead to more hardware documentation beeing released for ATI.

    5. Re:Don't really know.. by swv3752 · · Score: 1

      ATI, was releasing specs and even publishing OSS drivers to XFree86, and the morons at XFree86 spurned ATI. So ATI took thier highend opengl drivers, tweaked them sothat they would work on thier other cards and published them so that all Linux users would at least have something.

      --
      Just a Tuna in the Sea of Life
    6. Re:Don't really know.. by PhrostyMcByte · · Score: 1

      It's a bit of an odd buy, considering nVidia's chipsets contributed so much toward the rise of the Athlon and are still (afaik) the lead in performance chipset-wise. But I guess if they merge their teams (or at least make them consult eachother a lot) we could see some perf improvements.

      I wouldn't mind seeing a mini-gpu inside of the CPU dedicated to massive vector ops - it would certainly blow the pants off SSE2 for larger datasets (video/sound codecs?). Of course stuff can already use the GPU but I bet plain CPU instructions would make it a lot simpler!

    7. Re:Don't really know.. by sbrown123 · · Score: 3, Interesting

      Think multi-core CPU where one (or more) of the cores could serve as a graphics processor on demand. Oh, the power...

    8. Re:Don't really know.. by MrNemesis · · Score: 2, Insightful

      It's not just the nerds being more Linux minded - AMD has, to some extent, bet the farm on the K8 being the king of the server room, since the entire core was designed from the off to be highly scalable across multiple CPU's. ANd now we're seeing that most of the big advances (new "enterprise" sockets, K8L stuff) are going to benefit the servers before they benefit Joe Public.

      AMD knows that, whatever market share it has in the desktop arena, Linux is a major player in the HPC and 2P+ spaces and knows that Linux sysadmins won't tolerate buggy chipsets or flaky binary drivers that may end up being unsupported under kernel 2.8 or whatever. Hopefully AMD has the nouse to do an Intel and make their chipsets open spec across the board a la Intel, enabling excellent support under Linux and any other OS that happens to come along. I know for a fact that sometimes shoddy chipset support under Linux has been a turnoff for me in the past and I've lusted after some of Intel's chipsets on my own A64 systems.

      --
      Moderation Total: -1 Troll, +3 Goat
    9. Re:Don't really know.. by The+Raven · · Score: 1

      It's called a Cell processor.

      Well, maybe not quite. But the Cell is designed for the kind of grunt, parallel type work that graphics processors often do. Just remains to be seen if it works as well as IBM hopes.

      --
      "I will trust Google to 'do no evil' until the founders no longer run it." Hello Alphabet.
    10. Re:Don't really know.. by PhYrE2k2 · · Score: 1
      Better integration == better graphics/faster machines?


      Like when they started embedding SiS chipsets and graphic chipsets onto motherboards... Who needs stability when you have cheap hardware?

      -M
      --

      when you see the word 'Linux', drink!
    11. Re:Don't really know.. by Andy+Dodd · · Score: 1

      "There's a slight chance that AMD might be smart and release hardware specs for ATI cards, or make the drivers Free Software. If either of those things happen, this would be a very good thing, in my opinion."

      There's a good chance that they legally have no ability to do so, thanks to technologies that are licensed from other companies. (The example I can recall at the moment is S3 Texture Compression, support for which is the main "new" feature ATI's first binary drivers for Linux had over the previously fully functional open-source drivers.)

      --
      retrorocket.o not found, launch anyway?
    12. Re:Don't really know.. by babbling · · Score: 1

      So they leave that out of the Free Software driver they release. Big deal.

    13. Re:Don't really know.. by srock2588 · · Score: 1
      Oh, the power...
      Oh the heat! My god man...the HEAT!
      --
      Ehh...this is the life we chose.
    14. Re:Don't really know.. by Anonymous Coward · · Score: 0
      There's a slight chance that AMD might be smart and release hardware specs for ATI cards, or make the drivers Free Software. If either of those things happen, this would be a very good thing, in my opinion.


      My last three systems have been AMD with NVIDIA graphics, but I plan on building a Core 2 system with NVIDIA graphics in October or November. If AMD were to do this with ATI (release the specs) I would turn on a dime and by an Athlon X2 with ATI video in a heartbeat. ThI'm not holding my breath though.
    15. Re:Don't really know.. by scottnews · · Score: 1

      except that a CPU and a GPU are 2 entirely different processors

    16. Re:Don't really know.. by sbrown123 · · Score: 1

      Which usually gets hotter: a GPU or a CPU? Compare heat sinks sizes for a good guess. Dual core (and soon quad core) processors are already a reality. So I don't think heat would stop such an invention.

    17. Re:Don't really know.. by Andy+Dodd · · Score: 1

      You know what you get?

      You get a crippled piece of silicon that costs far more than an Intel GMA950, but has almost no performance benefits over aforementioned crippled chipset.

      There's a reason Intel's graphics chipsets have fully open-source drivers - It is because the chip is so crippled that it isn't subject to any of the intellectual property problems pervasive with higher performing chipsets. I would not be surprised if there is EXTENSIVE cross-licensing of technology in the graphics industry, where every company is using another company's technologies under license. S3TC is just the earliest (and most visible due to the Unreal Tournament 2003 + Linux + ATI fiasco) example.

      --
      retrorocket.o not found, launch anyway?
    18. Re:Don't really know.. by ChronosWS · · Score: 1

      Wow, awesome. Now when I want to upgrade my graphics processor, I have to get a whole new CPU. NOT.

    19. Re:Don't really know.. by prisoner-of-enigma · · Score: 1

      Think multi-core CPU where one (or more) of the cores could serve as a graphics processor on demand. Oh, the power...

      A general purpose CPU like the "Hammer" line is poorly suited to doing the kinds of things done by GPU's like NVIDIA's GeForce or ATi's Catalyst line. Graphics work requires a lot of vector work which is signficantly different than what a CPU is optimized for. In simple terms, vector works best when you have millions of tiny bits of data that all need the same kind of functions applied to them (i.e. pixels being lit, shaded, antialised, etc.), but vector is poor when you need to do general-purpose computations on other things (i.e. AI for 'bots and such).

      In short, what you're suggesting is not worthwhile. Now if you wanted to talk about putting a GPU or Physics-Processing Unit (PPU) in a CPU socket, connected via HyperTransport to main memory and the CPU, then you've got something. Unfortunately for you, AMD has already thought of that, and numerous companies are out there ready to produce custom FPGA's performing massively-vectorized calculations for industries like oil and gas exploration. Gaming can't be far behind if this trend takes off (which is a big if).

      --
      In the end they will lay their freedom at our feet and say to us, Make us your slaves, but feed us. - Fyodor Dostoyevsky
    20. Re:Don't really know.. by generationxyu · · Score: 1

      Wait... how?

      You say "They can really make a business-blow by releasing this, in the mind of open-source" -- but how exactly do you expect that to help them? I mean, I read slashdot. I like Linux. I'd like better support for high-end graphics technology under Linux. But that's maybe 3% of the market. The majority of the high-end graphics card market uses Windows, and wants performance. Period. If ATI is putting out better cards than the competition, they'll sell more cards. Period. Not by documenting their technology so nerds can play Tux Racer at 600 fps.

      --
      I mod down pyramid schemes in sigs.
    21. Re:Don't really know.. by rising_hope · · Score: 1

      I don't know that I would really agree with that. I think people who have the common sense to look at AMD as a viable cost-effective, performance orientated alternative to Intel also look at Linux as a cost-effective, sercurity orientated alternative to Windows. I run Windows... and Linux. They each serve a purpose. I also run AMD and Intel. My desktop is AMD, because it gave the best price/performance at the time of purchase. My laptop is Intel, because it gives the best performance/battery life at the time of purchase. If the tables flip, I'm apt to change. I have no loyalty to Intel or AMD specifically. I'm not sure it's that AMD-fans/nerds are more Linux-minded than Intel, but rather than people who haven't just gone with the default (Intel/Dell/Windows) are more apt to be Linux-minded.

    22. Re:Don't really know.. by rising_hope · · Score: 1

      Or better, that one of the GPU cores (SLI/Crossfire ensures they'll be more GPUs in the future...) could be used an additional CPU on demand. GPUs are MUCH better at math operations than CPUs, and could be great for supplimenting MMX type vector instructions.

    23. Re:Don't really know.. by srock2588 · · Score: 1

      Way to notice the Seinfeld reference, it's a joke...ass.

      --
      Ehh...this is the life we chose.
    24. Re:Don't really know.. by Anonymous Coward · · Score: 0

      "The AMD-fans/nerds are more linux-minded then Intel (IMHO), and AMD probably knows this. They can really make a business-blow by releasing this, in the mind of open-source."

      Yet Intel for years has gone way out of its way to release specs and open source drivers for their 3D chipsets, which work out of the box with Linux distributions, or *BSD. They release specs for their chipsets and controllers etc. etc. They contributed lots of money to gcc development to improve it for Itanium.

    25. Re:Don't really know.. by Jacek+Poplawski · · Score: 1

      There's a slight chance that AMD might be smart and release hardware specs for ATI cards, or make the drivers Free Software. If either of those things happen, this would be a very good thing, in my opinion.

      That would be not good but AWESOME, but why AMD would release specs for Radeons? Is AMD known for releasing documentation for their products?

    26. Re:Don't really know.. by Bert64 · · Score: 1

      By documenting the technology and open sourcing the drivers, performance and stability improvements in the drivers (which are what lets ATI down the most currently, on both windows and linux) will be forthcoming.

      Remember when matrox open sourced the drivers for some of their earlier cards? These cards quickly became the fastest 3d performers on linux, even beating out some theoretically faster hardware running on windows.

      --
      http://spamdecoy.net - free throwaway anonymous email - avoid spam!
    27. Re:Don't really know.. by Bert64 · · Score: 1

      Well, that was xfree86, and a lot of people had problems with the xfree86 developers...
      So how about xorg? are they any better? and would ATI work with them?

      --
      http://spamdecoy.net - free throwaway anonymous email - avoid spam!
    28. Re:Don't really know.. by Bert64 · · Score: 1

      CPUs and FPUs used to be entirely different chips too.

      --
      http://spamdecoy.net - free throwaway anonymous email - avoid spam!
  5. nVidia Counter Offer? by d3bruts1d · · Score: 1

    I don't suppose there is a chance for nVidia to make a counter offer is there? I know, wishful thinking on my part.

    nVidia: "No! Don't buy them! We're better, and we'll sell to you cheaper!"

    1. Re:nVidia Counter Offer? by Wudbaer · · Score: 2, Insightful

      Really great idea NOT. This would create more or less a monopoly for high-end graphic chips for NVidia. Who else is there ? Intel with their cheap chipset graphics ? Matrox ? A handful of far eastern companies that produce cheap and sucky low-end graphics products noone uses ? Nvidia buying ATI would be the worse thing that could happen for the consumer, even worse than the Intel quasi-monopoly of the dark years before the Athlon. As the history of Intel-vs.-AMD cleary shows competition not only is good for the consumer but in the end also for the companies which are required to innovate and improve their products and to keep themselves strong and vital.

    2. Re:nVidia Counter Offer? by fizzfaldt · · Score: 1

      I believe grandparent meant that nVidia would try to have AMD buy nVidia instead of ATI.
      'nVidia: "No! Don't buy them! We're better, and we'll sell to you cheaper!"'

  6. Linux Support ? by Anonymous Coward · · Score: 0

    This is a pretty obvious question, but hopefully it means ATi gets their shit together re: linux support . Getting any of their latest gear working under linux is pretty painful.

    1. Re:Linux Support ? by CowboyBob500 · · Score: 3, Informative

      Getting ANY of their stuff working under Linux is painful.

      I'm running OpenSuSE 10.1 on my Thinkpad R51 with a pretty standard ATI Mobility Radeon and can I get the ATI drivers working? Can I hell. Always "no device for screen" or some such error. So I'm stuck with the OSS drivers which although are great for 2D, don't perform well enough for anything other than TuxCart.

      On the other hand, the NVidia FX5900 in my desktop machine (also running OpenSuSE 10.1) was a breeze. Drop to run level 3, run installer, reboot, job done.

      Bob

    2. Re:Linux Support ? by Ash-Fox · · Score: 1
      I'm running OpenSuSE 10.1 on my Thinkpad R51 with a pretty standard ATI Mobility Radeon
      You should have 3d acceleration out of the box with the default stuff that comes with X.org (if you install FGLRX, it messes with your mesa libraries and won't give you 3d acceleration).

      So I'm stuck with the OSS drivers which although are great for 2D, don't perform well enough for anything other than TuxCart.
      I have 3d acceleration with the OSS drivers I have (they just won't get 3d acceleration if fglrx is installed). I have a Radeon Mobility M6 LY -- Pretty standard.
      --
      Change is certain; progress is not obligatory.
    3. Re:Linux Support ? by Anonymous Coward · · Score: 0

      reboot? what for? you can just load the module and startx.

    4. Re:Linux Support ? by Homology · · Score: 3, Interesting
      On the other hand, the NVidia FX5900 in my desktop machine (also running OpenSuSE 10.1) was a breeze. Drop to run level 3, run installer, reboot, job done.

      NVidia seems to make better blobs than ATI, but it is still a blob:

      Blobs are expedient. Many other open source operating
      systems cheerfully incorporate them; in fact their
      users demand them.
    5. Re:Linux Support ? by CowboyBob500 · · Score: 1

      To be honest, I really don't care in what form the driver comes, as long as it works.

      Bob

    6. Re:Linux Support ? by CowboyBob500 · · Score: 1

      FGLRX is long gone. But I still don't get a great frame rate in 3D. I get the acceleration, it's just not very good.

      Bob

    7. Re:Linux Support ? by Anonymous Coward · · Score: 0

      Re-boot? sorry that is an alien term to me?
      a quick "emerge nvidia-kernel && nvidia-glx" will get me the latest driver installed (although there is a migration to nvidia-drivers or something) and that can be done from a GUI, true it is still running X and thus the old driver is in use

      I can goto a VT and go "/etc/init.d/xdm stop" or in yr case you "init 3" BUT from there after driver is installed you do not need to reboot

      rmmod nvidia && modprobe nvidia

      will remove the old nvidia module from the kernel allowing the new one to be loaded
      Then a quick "/etc/init.d/xdm start" or "init 5" and you are back to the desktop no reboot
      Reboot only needed for kernel changes or major disk fixes

    8. Re:Linux Support ? by CowboyBob500 · · Score: 2, Interesting

      Typing "init 6" and then making a cup of tea for when my machine comes back up is far more relaxing.

      Not everything has to be done now, now, now. At least not in my world.

      Bob

    9. Re:Linux Support ? by Ash-Fox · · Score: 1
      FGLRX is long gone. But I still don't get a great frame rate in 3D. I get the acceleration, it's just not very good.
      Curious, do you get better performence under Windows?
      --
      Change is certain; progress is not obligatory.
    10. Re:Linux Support ? by heinousjay · · Score: 1

      That many extra steps to save one minute seems to miss the point just a touch.

      --
      Slashdot - where whining about luck is the new way to make the world you want.
    11. Re:Linux Support ? by CowboyBob500 · · Score: 1

      I don't have Windows installed anywhere, so I don't know.

      Bob

    12. Re:Linux Support ? by Anonymous Coward · · Score: 0

      ATI driver installation was easy under SuSE 10.1. I downloaded the driver fom ATI's site. Ran INIT 3. Ran the driver. Accepted the defaults, and then INIT 6. I don't know if INIT 5 would have worked, but I had nothing else critical running on the system. When it came back up, I tested with glxgears and got frame per second rates around 2500. All of the games I have tried work great. This is on a $200 walmart Micron computer that shipped with Xandros. Now it dual boots XP and SuSE 10.1. Games work well on both ends. I forget the model number of the Radeon I am running.

      The one drawback is that the xgl eye candy does not like the ATI cards. I can't get it to work.

    13. Re:Linux Support ? by CowboyBob500 · · Score: 1

      I think its because I've got a Mobility Radeon in a Thinkpad. It's unclear from the ATI website whether they are supported or not. Basically I followed the steps you've described as well. No dice...

      Bob

    14. Re:Linux Support ? by Dave2+Wickham · · Score: 2, Insightful

      Whilst I use the nVidia blob (actually I'm using a Matrox Millennium II at the moment, but I digress), the reason why people want open source drivers is so that if it doesn't work, they can make it work, rather than relying on nVidia to fix it.

      Another benefit would be that if nVidia's drivers were GPLd, they could be included with the Linux kernel and X(org|Free86) if they were to a high enough standard, completely eliminating the current issue of having to kill X to install the drivers, and reinstall with every kernel update; an open source driver would be far simpler to work with for all users.

      Whilst the blob is, IMO, better than nothing, I'd still much rather prefer good OSS drivers.

    15. Re:Linux Support ? by Tom · · Score: 1

      Getting ANY of their stuff working under Linux is painful.

      Not much better for windos, btw. - you have to manually patch their driver releases before they support many mobile versions of their cards. I'm not talking about binary patches to the core drivers - just the installer so that it freaking copies some files.

      Buying a notebook with ATI graphics? Worst mistake, ever.

      --
      Assorted stuff I do sometimes: Lemuria.org
    16. Re:Linux Support ? by Tim+C · · Score: 1

      But take away his uptime, and he has one less reason to feel superior to his Windows-using friends!

    17. Re:Linux Support ? by grim4593 · · Score: 1

      Maybe so, but his e-penis gets longer if he does it his way.

    18. Re:Linux Support ? by baadger · · Score: 1

      ...and in killing Xorg and, thereby your GUI and all X/GTK/QT apps, you lose almost everything you're doing anyway, it's like having a web server with 2 years uptime but apache hasnt been up for most of it.

    19. Re:Linux Support ? by bfree · · Score: 1

      You should probably be sharing the blame out betwen OpenSUSE and ATI. I spent many long hours trying to coax an ATI driver into working on various distros a few years ago (just to compare it's performance to the Free driver as I have r2*0 chips), then I discovered Kanotix where you can simply issue one Kanotix command and have it install the non-free driver perfectly for you (or upgrade it to the latest). No more worrying about the interactions of ATI's blob with your system as the distro team have taken care of it for you. Of course I only ever used it on throwaway test systems as I preferred to stick with the free drivers (and it's why I bought those machines/cards).

      Of course it would be nicer if people didn't feel the need to contaminate their systems with binary drivers but there do not seem to be an alternative if you want current card to give you anything like it's potential 3d performance. Time to start lobbying AMD to provide documentation for their next video chip (and even for past and current ones) and maybe even to try and organise an interested group to pay them to provide the documentation and support to create a good open driver (there must be more companies like the weather channel out there who will pay some cash to have good 3d support for a chip on linux).

      --

      Never underestimate the dark side of the Source

    20. Re:Linux Support ? by NotBorg · · Score: 1

      "Not everything has to be done now, now, now. At least not in my world."

      Call me young--I'm 30 and I'd like that reassurance--but how long should one wait? I personally don't want to wait until my hardware is obsolete to use it otherwise I'd just buy obsolete hardware to begin with.

      What can AMD and ATI bring to the table as a team? I find this whole line of discussion is interesting. But the questions that rack my impatient young mind are what will they bring and when will they bring it?

      Don't get me wrong, I am optimistic but at the same time my ATI card is going on three years. That's a long time when product life cycles--both hardware and software--are taken into the picture.

      A "cup of tea" sells reliable time tried products, but "Now, Now, Now" sells new products. Both are not without merit.

      --
      I want this account deleted.
    21. Re:Linux Support ? by CowboyBob500 · · Score: 1

      A "cup of tea" sells reliable time tried products, but "Now, Now, Now" sells new products. Both are not without merit.

      Again, not in my world. Well thought out and stable products interest me. Start-up time is irrelevent. As is running speed to a certain extent. And I'm definitely not interested in having the "latest and greatest".

      I'm only 33 by the way, so you've not got long to wait to learn patience ;-)

      Bob

    22. Re:Linux Support ? by the_maddman · · Score: 1

      Nvidia is the same way for mobile chipsets. Have to patch the .inf file to get it to install on my Compaq with it's GeForce4MX. And of course, once you install the desktop drivers, they get all confused as to what resolution the panel is. Laptop graphics are painful. Period.

    23. Re:Linux Support ? by permawired · · Score: 0

      I have a Dell Inspiron 6000 with an ATI Mobility X300 64MB. I installed Mandriva 2006 and EVERYTHING worked on the laptop with the exception of the SD card reader out of the box. I've used the video in dual head mode as well with no trouble. Also on another workstation that I have at home which has a Radeon 9800 Pro 128MB in it worked just fine with an out of the box Fedora Core 4 install. Again using dual head.

  7. could be good.. by Tokin84 · · Score: 5, Interesting

    this could be real good if AMD's acquisition of ATI allows them to produce full chipsets in the same fashion Intel has with its Centrino line. let the competition begin!

    also, not official yet, as government regulatory bodies need to approve it.

    --
    Most human beings have an almost infinite capacity for taking things for granted. - Aldous Huxley
    1. Re:could be good.. by plague3106 · · Score: 1

      If AMD really 'owns' ATI, I would think they could simply dictate that.

    2. Re:could be good.. by Clith · · Score: 1
      also, not official yet, as government regulatory bodies need to approve it.
      Since this acquisition straddles two countries, I wonder how all the details work out?
      --
      [ReidNews]
    3. Re:could be good.. by Jeff+DeMaagd · · Score: 1

      AMD can and do make chipsets, but unfortunately, they have been behind for some time. For a while, AMD had the only ones for their multi-processor chips.

  8. Should we welcome our new overlords? :D by sihker · · Score: 1

    ATI linux drivers for AMD?

    1. Re:Should we welcome our new overlords? :D by Compholio · · Score: 2

      ATI linux drivers for AMD?

      Dear AMD:

      We, your faithful processor purchasers (yes, we have many), have long been forced to buy nVidia hardware because of ATI's poor quality drivers under Linux. Please work the same magic you did with the AMD64 and give us something we can be proud of.

      The undersigned.

  9. Probably Not Going To Happen... by FinchWorld · · Score: 5, Funny

    ... But hopefully they'll kick the ATI driver team up the arse and get a decent set of drivers out (for Windows and Linux).

    --
    "I may be full of crap about this game, and I may be wrong, and that's fine." -Jack Thompson
    1. Re:Probably Not Going To Happen... by d3bruts1d · · Score: 1

      ...or AMD could turn driver development over to the ATI employees, and just totally FUXOR the Linux crowd.

    2. Re:Probably Not Going To Happen... by morgan_greywolf · · Score: 1
      ... But hopefully they'll kick the ATI driver team up the arse and get a decent set of drivers out (for Windows and Linux).


      Modded funny? I don't think so.

      I think AMD knows that the Linux crowd is amongst their biggest fans. Let's face it, geeks like the underdog, especially when they produce competitive or even in some cases superior products than the market leaders. That's one of the reason's for Apple's recent successes on the Macintosh platform. Seriously.

      Maybe AMD will make their ATI division really appeal to us and open source their drivers. If they did, the crucial thing is that they will eat NVidia for lunch in the geek market.

      People on /. are always downplaying the geek market, but we're very affluent... especially when companies realize that we affect the buying decisions of everyone around us. When we recommend a product, it goes.
    3. Re:Probably Not Going To Happen... by Ilgaz · · Score: 1

      Don't forget OS X. Game companies get flamed because of their driver bugs.

      It is consumers fault too, they should report problems to ATI instead of flamebaiting game support boards etc. If a game crashes or displays horrible stuff, fps, one thing comes to average end users mind "It is buggy/broken".

    4. Re:Probably Not Going To Happen... by 0racle · · Score: 1

      They also know those fans aren't that large of a market. There is very little business reason to open up nVidia's or ATI's drivers. Think about it, the only reason anyone here comes up with is they might buy a card in the future if they did. Might, years from now. when their current 1mb card from 10 years ago blows up.

      --
      "I use a Mac because I'm just better than you are."
    5. Re:Probably Not Going To Happen... by morgan_greywolf · · Score: 1
      They also know those fans aren't that large of a market.


      You're kidding, right? I wouldn't be so sure.

      According to this Washington Post article:

      . With some analysts estimating the Japanese geek market to be worth as much as $19 billion a year, companies are jostling to cash in.

      This telephony online article is 4 years old, but believe me, the love is still there.

      This CRN article underscores the current hotness of the geek market.

      Look at all the advertising on Slashdot, Linux.com, Digg, and other geek hangouts...believe me, marketeers care about the geek market, and if AMD has any smarts at all, so do they.

    6. Re:Probably Not Going To Happen... by 0racle · · Score: 1

      Define geek market. Linux users make up a small amount of the purchases in it, most comes from hardcore gamers. They don't care about open drivers, they want raw performance. If you're going to make a business selling to the 'geek market' pandering to Linux users is going to keep you small, the money is coming from retarded gamers who truly believe that everything becomes obsolete in less then 6 months.

      --
      "I use a Mac because I'm just better than you are."
    7. Re:Probably Not Going To Happen... by mrchaotica · · Score: 1
      Define geek market.

      The market of people who give all the non-geeks buying advice. As it is, I advise people to steer clear of ATi. I really hope that will change, though.

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    8. Re:Probably Not Going To Happen... by 0racle · · Score: 1

      Oh you mean the kid next door that doesn't really know anything. Since this thread started on thinking that ATI should open it's drivers to pander to Linux enthusiasts, on those grounds shouldn't you also be advising people to steer clear of nVidia.

      --
      "I use a Mac because I'm just better than you are."
  10. Maybe by neuromancer2701 · · Score: 2


    System on a chip or at least integrated GPU and CPU cool.
    I just wish it was Nvidia.

    --
    "If you like Battlestar Galactica, you're probably a huge nerd." -Stephen Colbert
    1. Re:Maybe by Homology · · Score: 1, Flamebait

      > System on a chip or at least integrated GPU and CPU cool.
      > I just wish it was Nvidia.

      Why are posters so fond of the anti-open source hardware vendor
      NVidia? This is a company that refuses to release _any_ hardware
      documentation at all for any of its chipsets!

    2. Re:Maybe by jank1887 · · Score: 1

      let me rephrase that for you: System on a chip or at least integrated GPU and CPU. VERY NOT COOL. In fact, likely to melt a hole through your motherboard. Unless they supply stock liquid cooling, or cut power to both Units by 50% at the same time. Just too much heat in too little real estate. You need the separate fans currently in place. One over both wouldn't cut it. Integrated might save you a few watts if they could share some resources, but probably not too much.

    3. Re:Maybe by SchwarzeReiter · · Score: 1
      System on a chip or at least integrated GPU and CPU cool.
      You can get that from STI, it is called Cell.
    4. Re:Maybe by Linker3000 · · Score: 5, Funny

      "System on a chip or at least integrated GPU and CPU cool."

      A die holding an AMD core and an ATI GPU may be 'neat', 'fab', 'brill' or even 'ace' - but 'cool' - I think not!

      --
      AT&ROFLMAO
    5. Re:Maybe by Wdomburg · · Score: 1

      Via manages it with a TDP of 6-10W, depending on clockspeed.

    6. Re:Maybe by Rykky · · Score: 1

      You cant call NVidia anti-open source. I've been a happy NVidia customer for many years and I've always loved how they Support my Open Source Operating System. Can you point out the buisness case or value for NVidia to open up their hardware or drivers ?

    7. Re:Maybe by rabbot · · Score: 1

      Perhaps it is because Nvidia designs better cards and certainly has better drivers.

      Why would any rational person shun a company simply because they don't release documentation for their chipsets? The world doesn't revolve around open source.

    8. Re:Maybe by Giant+Ape+Skeleton · · Score: 4, Insightful
      Why are posters so fond of the anti-open source hardware vendor NVidia?

      Probably because most Slashdotters are not driver hackers nor OSS purists, they are developers, gamers, and power users -- and Nvidia's hardware (and driver support for the hardware) is phenomenal.

      Your gripe is not baseless, though: would it kill Nvidia to open up a bit? Perhaps the renewed competition will encourage them to do so, although it's equally likely that they will take the opposite tack and circle their wagons ever more tightly. As long as they provide excellent binary drivers for Linux, I doubt that they will feel much incentive to go Open Source...

      --
      The difference between stupidity and genius is that genius has its limits.
    9. Re:Maybe by Anonymous Coward · · Score: 0

      Yeah but AT LEAST, drivers for linux from nvidia doesn't suck.
      And nvidia is friendly to some unix's : we have IA64, x86 and x86-64 drivers for linux, and some drivers for FreeBSD and Solaris.

      ATi has some shitty, PISS POOR x86 and x86-64 for linux and that's all.

    10. Re:Maybe by shrubya · · Score: 1

      You are mistaking loudness for consensus. Surprising as it may appear, a majority (or at least a very strong plurality) of Slashdotters run Windows as their primary OS. And *nix pragmatists are comparable in number to your *nix idealist faction.

      I'm glad there are folks like you, willing to boycott hardware for the sake of OSS purity. I hope someday Nvidia capitulates. But unless they're draize-testing babies or dumping toxic waste in the rain forest, I'm using whatever best meets my needs.

    11. Re:Maybe by strider44 · · Score: 1

      I think it's because they regard the high quality of both hardware and support as more important than external documentation.

    12. Re:Maybe by GeeksHaveFeelings · · Score: 1

      Sure, nvidia doesn't offer an OSS solution, like releasing the interface specs so a OSS driver can be written, but at least it makes a good effort in making a proprietary drivers. ATI's drivers are crap, and they don't offer an OSS solution. And some of us are gamers, and we just don't care whether it's open source or not, since the nvidia proprietary drivers do provide better performance than ATI's.

    13. Re:Maybe by mungtor · · Score: 4, Interesting

      nVidia CAN'T open up their drivers IIRC. nVidia was founded by a bunch of SGI engineers, and once they started producing products the folks over at SGI found some of their technology in the nVidia products. As part of the settlement, nVidia can't release the code since I think they had to license it from SGI in the end.

      I could swear that's they way that it is, but I can't find any definitive reference to the settlement.

    14. Re:Maybe by Tom · · Score: 4, Informative

      Why are posters so fond of the anti-open source hardware vendor NVidia?

      Because they've supported Linux with binary drivers for a long time, and their drivers work.

      ATI is months behind, and half of the time the drivers are too buggy to actually use.

      Philosophy of openness aside, that's an important difference.

      --
      Assorted stuff I do sometimes: Lemuria.org
    15. Re:Maybe by Kjella · · Score: 1

      Well, from what I've understood the newest ATI cards aren't supported at all, closed source, open source, 2D, 3D or otherwise. How pathetic is that even for a dual boot machine? While nVidia won't release their own source, they are making Linux a damn lot more usable for both 2D and 3D. For many people the choice is not tainted vs untainted kernel - it's tainted kernel or no Linux at all. With all the battles you have to fight to use Linux for everything, you rarely fight against what already works. So there are a few open source drivers for ATI, are they impressive? Last I heard, no. Get ATI to give up all the specs, create some kick-ass drivers and put it to nVidia "Why can't we do that with your cards?" As it is, do you think they see much of a point in OSS drivers?

      --
      Live today, because you never know what tomorrow brings
    16. Re:Maybe by Richard_J_N · · Score: 1

      Indeed. Nvidia's drivers are a royal pain to install, but at least they are pretty good when running. That said, as soon as anyone makes a
      card with proper GPL drivers that can do dual-head,DVI,3200x1600, I shall switch. I like 3D accel, but I'd be quite happy with an unaccelerated frame-buffer - just as long as is stable. The OpenGraphics project is trying....

    17. Re:Maybe by Slashcrap · · Score: 3, Funny

      I like 3D accel, but I'd be quite happy with an unaccelerated frame-buffer - just as long as is stable.

      Go into your xorg.conf and change the "Driver" line to "vga" or "vesa".

      You now have an unaccelerated frame-buffer display that will be as stable as the day is long. It will also suck donkey cock, but it is exactly what you just asked for.

    18. Re:Maybe by Richard_J_N · · Score: 1

      Yes - but I have a dual monitor setup. 2 x 1600x1200 LCDs, both connected via DVI to the same card.
      The VGA/VESA drivers make them both display the same thing! So does the xorg "nv" driver.

    19. Re:Maybe by neuromancer2701 · · Score: 1

      To my knowledge the Cell does not have the GPU on the same die, Nvidia made a separate chip.

      --
      "If you like Battlestar Galactica, you're probably a huge nerd." -Stephen Colbert
    20. Re:Maybe by mark-t · · Score: 2, Interesting

      Isn't SGI in the process of folding? What happens to the contract once SGI is gone?

    21. Re:Maybe by SchwarzeReiter · · Score: 1

      The Cell has a couple of vector units inside, they can handle graphics operations. The NVidia chip just renders the display lists into the graphics ram, so I dont think it is a GPU in the sense we use the word, like the chips are in the PC accelerator cards. For example I think the shader operations will run on the Cell, and not on the RSX, so how I understand the things, the RSX will be something similar to the GPUs before the original Geforce, without T&L.

    22. Re:Maybe by Homology · · Score: 1
      nVidia CAN'T open up their drivers IIRC. nVidia was founded by a bunch of SGI engineers, and once they started producing products the folks over at SGI found some of their technology in the nVidia products. As part of the settlement, nVidia can't release the code since I think they had to license it from SGI in the end.

      I'm talking about hardware specs, NOT code. nVidia does not release docs for ANY of its hardware! What is so secret about a NIC that thay cannot release the docs?

    23. Re:Maybe by Anonymous Coward · · Score: 0

      I could have sworn I read a story on Slashdot a month or 2 ago where an insider familiar with the situation speaking on the condition of anonymity and said both ATI and nVidia are violating each others patents. Opening the driver documentation up would be the equivalent of opening a pandora's box of litigation for either of them. Since both are at fault, though they are not sure who is more at fault, the suits have decided to not be the first to invite litigation. If the other keeps their dirty secret about the patents they are violating they can't be sued, but will sue the other. Maybe the story was inaccurate, but it was certainly interesting.

    24. Re:Maybe by drinkypoo · · Score: 2, Funny

      Well, if that's the only thing stopping them, we should be in luck. In a month or two they should be able to pick up SGI at the flea market for five bucks "but only if you take that stack of documentation, too."

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    25. Re:Maybe by BlueLightning · · Score: 1

      I would assume SGI will need to sell its assets to cover debts, so someone else will end up with ownership of the IP.

    26. Re:Maybe by Al+Dimond · · Score: 1

      I don't know about sucking donkey cock. My old-ish laptop has some random GPU whose X drivers have major stability issues. So I use the vesa drivers and I've never found myself desiring more performance when switching desktops in X or shuffling windows around. A P3 with sufficient memory (the laptop has 192MB) is plenty capable of handling the load.

      I also have a P3 desktop box with just 64MB RAM that's currently running Plan 9, but used to run FreeBSD with the nVidia binary drivers. In that case the drivers made a big difference for every-day window management because I was running at a high resolution without much RAM. Of course, running at low resolution isn't that much of a hindrance thanks to xft (and the fact that programs old enough not to support xft generally work much better at low resolutions than high). I notice much less of a difference between my 1600x1200 desktop screen and my 800x600 laptop screen than you would think.

  11. I'd say I told you so.... by Groo+Wanderer · · Score: 5, Funny
    1. Re:I'd say I told you so.... by Anonymous Coward · · Score: 0

      Pathetic, it's like the retard shooting baskets all day long and by dumb luck on finally manages to have one go in.

      And then wants everyone to cheer...

      What is so sad about theinquirer is it has such a joke of a reputation that people no longer follow links to stories there in discussion boards because it is so obvious that anything being linked to at the site is so obviously wrong it doesn't even warrant the board's time debunking whatever story is being linked to.

    2. Re:I'd say I told you so.... by teaeg · · Score: 1

      I love "The Inquirer"!!!

      --
      A disgruntled economist
    3. Re:I'd say I told you so.... by Ilgaz · · Score: 1

      I can imagine how your mailbox after that article "printed" on site, you deserve a good "I am not bragging but.." article :)

    4. Re:I'd say I told you so.... by bagsc · · Score: 1

      I love Hector Ruiz, and AMD has had excellent execution since he's taken the helm, but I've got to wonder what he's thinking here. If he needs better integrated chipsets, you can enter a long term contract, and not hurt ATi's Intel business, or AMD's nVidia business as much. He's done a good job getting AMD down to a real CPU business, spinning off Spansion, and getting ready to ditch Geode. Why enter three new markets? (Chipsets, GPU, Handhelds)

      I think a reason is that this is a rebuilding year for AMD, and today begins the bloodletting of the price war. AMD more likely needs the cashflow and manufacturing licenses to survive an Intel assault in the next four quarters. Look up 'big bath' theory...

      I think Charlie's got it right - PCIE will be replaced with GPUs on motherboard, a la the HT2 multi socket idea. Whatever the reason, they need a real paradigm breaker, and to execute it well to make this hardball (nearly backstabbing nVidia) move look smart. I think the HT2 is just that, IF the integration of ATi's people works.

      This discomfort plus Conroe performance specs means that for the first time in seven years I'm pricing Intel parts.

      --
      http://www.accountkiller.com/removal-requested
  12. AMD & ATI by digitaldc · · Score: 5, Funny

    AMD combines with ATI and has announced a new name for their company:

    DAAMIT!

    --
    He who knows best knows how little he knows. - Thomas Jefferson
    1. Re:AMD & ATI by Anonymous Coward · · Score: 0

      this joke is getting better and better ;)

    2. Re:AMD & ATI by mgblst · · Score: 1

      This is funny, real funny. Better than slashdot funny. This is YouTube funny - can you make a video version?

    3. Re:AMD & ATI by digitaldc · · Score: 1

      Never saw that thread before, but alike minds think great!

      --
      He who knows best knows how little he knows. - Thomas Jefferson
    4. Re:AMD & ATI by Anonymous Coward · · Score: 0

      alike minds think great

      -or- Idiots seldom differ..

  13. Our Plan for Take Over by Anonymous Coward · · Score: 1, Funny

    1) Adobe buys Macromedia
    2) AMD buys ATI
    3) Intel buys nVidia
    4) We buy Adobe, AMD and Intel

    - Microsoft

    1. Re:Our Plan for Take Over by AkumaKuruma · · Score: 1

      Intel is not buying nVidia. theres too much overlap of products to warrant it. theyboth already make controller chipsets and graphics chips. Intel already leads as the #1 supplier of GPUs. its just theyre not gaming chips.Intel is already having to downsize the company a bit and looking at selling sections of its other departments off.

    2. Re:Our Plan for Take Over by ActiveMatx · · Score: 1

      Agreed. It is more likley that Nvidia were to buy Intel! (which isn't too far out there).

    3. Re:Our Plan for Take Over by mangaskahn · · Score: 1

      You left off a line: 5)Profit!

      --
      Really, I'm not out to destroy Microsoft. That will just be a completely unintentional side effect.--Linus Torvalds
    4. Re:Our Plan for Take Over by Anonymous Coward · · Score: 0

      Now i cant have a Intel E6700 with an ATI crossfire machine

    5. Re:Our Plan for Take Over by Wdomburg · · Score: 1

      A company with a market cap of $6B and revenue of $500M last quarter buying a company with a market cap of $100B and revenue of $8B last quarter "isn't too far out there"?

  14. This is a very good thing. by NXprime · · Score: 5, Interesting

    http://www.theinquirer.net/default.aspx?article=33 219
    *head asploded*

    I'm getting the 'gist' of why this transaction needs to happen. AMD needs GPU functionality on the CPU. I think everyone kinda expected that to happen at some point. The Inq. then takes a left turn in the plot and mentions 'mini-cores' which are multi-cores with massive amount of threads. Sort of but not really like Intels' hyperthreading times 32x. Shitloads of threads.


    Bottom line?

    ATI will work on AMD's new cores. I don't know if they'll work on something that'll plug into a PCIe slot still like nVidia.

    nVidia will still be around making graphic cards for AMD. Just won't necessarily be anything remotely similar to what's out on the stores today. AMD doesn't like closed technology like Intel does. So it'll be an open platform still which is a 'good thing' (tm).

    Forget about GPU's and chipsets. The main innovation has to come from these new GCPU's.

    ATI was going to lose its Intel chipset business anyway with or without this takeover. So no big loss here.

    Intel has about a year lead on this tech and probably be first out to market with it.

    CPU cores change radically every 5 years or so. With GCPU's, think more in terms of GPU's and radical changes every year to 18 months. Crazy shit.

    Plenty of space at FAB 36 to build the new cores and the recently announced plant they are building in New York. So no more costly production runs in Taiwan.

    If AMD didn't do this, they'd be out of business in 5 years. Period.

    1. Re:This is a very good thing. by Anonymous Coward · · Score: 0

      If i have to buy a new processor for every video card upgrade this will not be a good thing.

    2. Re:This is a very good thing. by Groo+Wanderer · · Score: 3, Interesting

      Think of it more as adding instruction sets to the CPU, not adding a GPU to the CPU. MMX, not embedded graphics.

                  -Charlie (the author of the Inq article)

    3. Re:This is a very good thing. by plague3106 · · Score: 1

      AMD needs GPU functionality on the CPU. I think everyone kinda expected that to happen at some point.

      Er, I don't, for the same reason I don't buy MBs with integratd graphics; in 3 months, when the next twice as fast GPU comes out, I'd like the option of upgrading it. A processor seems to have a longer life than a graphics card.

    4. Re:This is a very good thing. by fbjon · · Score: 1
      AMD needs GPU functionality on the CPU
      Eventually we'll all probably just have huge processor packages that we plug expansion cards into, instead of a motherboard.
      --
      True confidence comes not from realising you are as good as your peers, but that your peers are as bad as you are.
    5. Re:This is a very good thing. by Anonymous Coward · · Score: 0

      AMD doesn't like closed technology like Intel does. So it'll be an open platform still which is a 'good thing' (tm).

      AMD is just as full-on for closed technology as Intel. They are working hard on their Trusted Computing chipsets (TPM inside the CPU)... AKA DRM in hardware -- google for "presidio".

    6. Re:This is a very good thing. by mnmn · · Score: 1

      This is where the technology is headed. I'm sure we'll see CPUs with an integrated GPU sans memory along with CPUs without GPUs for cheaper machines. We've already seen the Northbridge integrated onto the AMD CPUs.

      I would like to see the first 128MB of RAM into the CPU housing as well as the GPU and the minimal southbridge. This should bring motherboard prices lower at the cost of a higher CPU, the overall cost should still be lower. Even better, it should allow for some serious speeds.

      At the minimum, the GPU should move into the CPU housing but not on the same die (that'll be expensive) along with 128MB of RAM and 4 cores.

      --
      "Give orange me give eat orange me eat orange give me eat orange give me you." -Nim Chimpsky
    7. Re:This is a very good thing. by A+Brand+of+Fire · · Score: 2, Interesting

      An interesting hypothesis that came to mind during and after the confirmed speculation and in light of AMD's announced 4x4 platform: plug-in GPU modules on the motherboard. With the way 4x4 works, you would be able to dedicate determinable and upgradeable RAM to the GPU. And since ATI and nVidia have both been working integrating a PPU core in future GPUs, there are interesting possibilities on the horizon.

      Having a bank of RAM slots on the motherboard in dedication to an socketed GPU has its drawbacks, I'm sure, but I kind of like the idea of dropping in a new GPU chip with a mobo-mounted cooler.

      As to how it'll affect AMD's and ATI's business model in the future, and their very futures for that matter, we'll have to wait and see how it pans out. I think it's a good move on both companies' parts, despite the immense risk involved. To quote the sixty-second Rule of Acquisition, "The riskier the road, the greater the profit."

      --
      [End of Line]
    8. Re:This is a very good thing. by CastrTroy · · Score: 1

      It really depends on how often you plan on buying a new video card. Myself, I buy a new video card about as often as I buy a new processor/mainboard/ram/everything else. However I shy away from onboard video because it sucks pretty bad. If they could through together a decent onchip GPU it would have super low latency, and could have a lot of extra advantages.

      --

      Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
    9. Re:This is a very good thing. by NXprime · · Score: 2, Insightful

      Here's what I don't get. If they do that, how do you upgrade the memory bus bandwidth so that it's futureproof to a degree? Memory on graphics cards changes all the time. It's not just a GPU and Memory. It's everything in-between as well. Power voltages... ect.

    10. Re:This is a very good thing. by NXprime · · Score: 1

      Not the same thing. What companies (legit that is) does this prevent from being apart of system componets? I'm talking hardware, not software. Besides there'll be a hack to bypass that HDCP stuff anyway. Most likely the printscreen key, a magic marker, or something like that :)

    11. Re:This is a very good thing. by dreamchaser · · Score: 2, Interesting

      Not at all a good idea, nor where things will be heading except in the embedded market perhaps. Not only would it reduce yields enormously because of the larger die size, but it would also put two points of failure into one chip, AND make it much harder to upgrade just one component.

      I could see this perhaps in the mobile/embedded market, but not in the server/workstation space. At least not for a LONG time. It's just not a good idea.

    12. Re:This is a very good thing. by SenseiLeNoir · · Score: 1

      Its to do with the Hyper Transport technology that AMD already has. HT is already a system bus of sorts, which can have modules attached to it.

      --
      Have a nice day!
    13. Re:This is a very good thing. by NXprime · · Score: 1

      No if HT works up to 2ghz and a new GPU comes out and needs 3ghz, well... see the point? it's about point to point speed with HT being the bottleneck for upgrades.

    14. Re:This is a very good thing. by plague3106 · · Score: 1

      Personally, I don't buy them very often, but I like having the option. Typically though, I go through two or three graphics boards before I find that I need to upgrade my processor as well.

      I had an AMD 1800+, and started off with an ABIT GeForce4. Later (few years ago), I got an FX5700. It wasn't until just recently I moved to an X2 3400+ (I think).

    15. Re:This is a very good thing. by Anonymous Coward · · Score: 0

      Yes... it is the same thing. Besides, if the software is changed, then the hardware no longer trusts it and refuses it access to trusted data... or may not run it at all.

    16. Re:This is a very good thing. by eXFeLoN · · Score: 1

      Costly production in Taiwan? Is this new? I'm under the impression that's why the production is in Taiwan, because of cheap labor manual or otherwise...

      --
      My other sig is a knife wound.
    17. Re:This is a very good thing. by Dan+Ost · · Score: 1

      No problem. The GPU uses a clock multiplier internally but talks 2ghz over HT when
      it needs to send something off-chip. Piece of cake (unless there's something I'm
      missing about how HT works).

      --

      *sigh* back to work...
    18. Re:This is a very good thing. by mnmn · · Score: 1

      Oh no I didnt mean on the same die. Heck it might even be in a different housing (like the FPU chip placed on top of the CPU or like the L2 cache on slot1 cartridges). The thing is instead of just standardizing a bus (AGP/PCIX) and waiting a year for people to follow and motherboard manufacturers to get it right, the CPU makers will just connect the GPU to the CPU with hyperwhatever, and clock it as far as it goes. Lower yields like celeron/duron might have no GPUs or slower bus speeds etc like the 533MHz 800MHz FSB difference. PCIX is damn fast. But we'll hit that wall soon, and someone will have to come up with a newer standard there. In fact PCIX is too fast for many applications and will soon be too slow for graphics, especially if you add a physics chip there as well.

      Any other reasons why its a bad idea?

      --
      "Give orange me give eat orange me eat orange give me eat orange give me you." -Nim Chimpsky
    19. Re:This is a very good thing. by ysachlandil · · Score: 1

      Or maybe ATI will make a GPU with built-in memory controller (like xbox360) and HyperTransport bus. No more PCIe or whatever, just a simple and fast HT pipe to the CPU. Puts the GPU closer to main memory (where it needs to be) and looks just like normal memory.

      --Blerik

    20. Re:This is a very good thing. by dnoyeb · · Score: 1

      Why not integrate the GPU with the chipset? Like we do today...

      The GPU with the CPU would be far too much heat in one place. Furthermore think of the product line complexity. I can imagine the name of a GCPU XP97005600+2. Insane.

    21. Re:This is a very good thing. by drinkypoo · · Score: 1
      Er, I don't, for the same reason I don't buy MBs with integratd graphics; in 3 months, when the next twice as fast GPU comes out, I'd like the option of upgrading it. A processor seems to have a longer life than a graphics card.

      In practice, it would probably complicate things all out of proportion to the benefits, but it's quite possible that we'll see a pluggable GPU architecture like we have for CPUs now, with a motherboard with sockets for both. On the other hand, for this to not be a total crapfest for the consumer, it will be necessary for CPU sockets to stabilize and reduce in number, and for GPU sockets to have some longetivity.

      Still, there's no reason that we can't have mainboards with CPU and GPU sockets, and CPU and GPU memory banks, and we could be communicating between CPU and GPU with hypertransport.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    22. Re:This is a very good thing. by labratuk · · Score: 1
      AMD doesn't like closed technology like Intel does.

      Intel's graphics chips have the most comprehensive DRI drivers out there. Want to know why? Intel released helped write them and released full docs.
      --
      Malike Bamiyi wanted my assistance.
    23. Re:This is a very good thing. by obeythefist · · Score: 1

      MMX isn't very good... sure we all have it and it's probably a mainstay of CPU instruction sets but it didn't really achieve what Intel wanted it to, which is why we have seperate GPU's. Specialisation is king, and this merger doesn't change it. A CPU is a great general purpose processing engine, but throwing GPU instructions at it is more like putting lipstick on a pig in terms of graphics processing performance.

      Now here's an idea... AMD have proven to be quite good at slapping CPU cores on a chip. The real estate of a GPU isn't wildly different from that of a CPU. What kind of performance would you get integrating a GPU and a CPU on the same die with hefty on-die interconnects? (Not to mention a presumably integrated memory hub for both of those cores, AMD style).

      Now all you need is an ultra-fast memory interface (graphics memory is always speed hungry) and you can do away with graphics cards altogether.

      What about SLI and dual graphics solutions? AMD is already on 4 cores in the very near future.

      Imagine that!

      --
      I am government man, come from the government. The government has sent me. -- G.I.R.
    24. Re:This is a very good thing. by acklaiber · · Score: 1

      >> If AMD didn't do this, they'd be out of business in 5 years. Period.

      How about the other way around? If ATI didn't do this, they would be out of business in (N) years. To quote a 2004 interview with Glenn Henry of Centaur-now-Via: "It's clear that the processor is the black hole, and that all silicon is going to fall into it at some point. [...] Those who can do a processor can control their system destiny, and those who don't will end up totally at the mercy of other people, who can shut them out of business right away."

      Under this perspective, this acquisition may well be more in ATI's interest than AMD's. Admittedly, it's a mystery why ATI would get paid good money for it. Another more cynical view is that this is a case of two second-tier companies thinking that by merely combining forces they can bet the incumbents. IMHO, dilution of focus is more like it...

    25. Re:This is a very good thing. by Anonymous Coward · · Score: 0

      Look what the fastest GPUs do and what makes them so fast: 48 Pixels per clock

      Next, look what turn Sun, Intel and maybe AMD are taking, towards CPUs that can run more than32 concurrently running threads.

      Now, guess how many concurrently running threads our CPUs can run today? It doesn't matter how good or bad MMX or 3DNow are without abilities to run massively parallel tasks.

  15. Man... by GundamFan · · Score: 1

    This will likely put a functonal end to the Intel processor + Intel chipset + ATI Video card systems I like to build.

    It wouldn't be so bad if every nVidia based product I have ever tried to use hadn't been DOA.

    Well... at least I can still stick with Intel chipsets... there is no way I am using a third party northbridge/southbridge I don't care if I can't use SLI.

    --
    I don't give a damn for a man that can only spell a word one way.
    Mark Twain
    1. Re:Man... by plague3106 · · Score: 1

      It wouldn't be so bad if every nVidia based product I have ever tried to use hadn't been DOA.

      You must have very bad luck. Or perhaps you're buying from ultra cheap board manufacturers? All the NVidia stuff I've had has worked great, with the exception of one card. It melted down trying to play Doom3, but I was able to get it replaced free of charge since it was new.

    2. Re:Man... by GundamFan · · Score: 1

      Yeah you are probably right... At least with ATI you can buy the name brand cards (built by ATI) and know that it will have good warenty support and a low chance of defects.

      Is there any one (or two) nVidia based card manufacturers that you trust? I, as you can see, have been bad at picking winners so far.

      --
      I don't give a damn for a man that can only spell a word one way.
      Mark Twain
    3. Re:Man... by bohemian72 · · Score: 1

      I've had good luck with ASUS and ever Chaintech and Gigabyte.

      --
      The greatest thing you'll ever learn is just to love and be loved in return.
    4. Re:Man... by akepa · · Score: 1

      "It wouldn't be so bad if every nVidia based product I have ever tried to use hadn't been DOA."

      You're not alone. Earlier this year I ordered 3 various nVidia cards in a row, and they all either were DOA or gave me BSOD's every 5-10 minutes. Finally bought an ATI X1800 GTO, and I couldn't be happier with it. I don't care if the Linux drivers suck -

    5. Re:Man... by Anonymous Coward · · Score: 0

      You're a maroon. Please die. Thank you.

    6. Re:Man... by Anonymous Coward · · Score: 0

      When my Gigabyte mobo died I replaced it with an MSI one as my other Comp. Sci. housemates recommended them.

    7. Re:Man... by plague3106 · · Score: 1

      ASUS or ChainTech have been pretty reliable for me (although the ChainTech was the one that melted, its replacement is still running now years later). Also, I had an ABIT card that seemed pretty decent.

    8. Re:Man... by jp10558 · · Score: 1

      XFX and ASUS are good in my experiance.

      --
      Opera, Proxomitron-Grypen,GPG 0x0A1C6EE3
    9. Re:Man... by arashi+no+garou · · Score: 1

      To the folks above having problems with your nVidia cards being DOA/giving BSODs...I'd be willing to bet it's not the card but your PSU. I recently got a GeForce 6600GT which has a power port on it requiring a drive-style power connector to be plugged in. This card is a power hog; my "measly" 450 watt, highly recommended power supply couldn't handle the card without lockups and such. At the minimum I'd say go for a 550 watt PSU with any card from the 6xxx or 7xxx series.

    10. Re:Man... by mrxak · · Score: 1

      I had one nVidia card explode after a few months, and then its replacement was DOA. I bought an ATI that supposedly had about equal performance for cheaper than the nVidia card, and it's been performing far better. I don't think I'll be getting another nVidia card for a while, though I suppose after this AMD-ATI merger, I'll have to see what happens. In a few years when I make my next GPU purchase, things may be radically different with both companies.

    11. Re:Man... by Anonymous Coward · · Score: 0

      You must come from a parallel universe, intel + intel + ati has been just so absolutely pants compared to amd + nv + nv..

      You want an intel rig, go buy a fscking Dell and never ever open the case. You want a souped up macho machine held togheter by duct tape and cooled by liquid helium, you go buy amd and nvidia. That's how it's been basically since AMDs Socket A. YMMV, but for enthusiast computer builders the combo has been a gem, both in terms of quality, cost and performance.

    12. Re:Man... by namityadav · · Score: 1

      ATI doesn't build it's cards .. Sapphire does. ATI just pastes the 'ATI' name on the cards.

      eVGA and XFX build very stable nVidia cards. In fact both of them have life-time warranty on the cards and both (Specially eVGA) are considered the Google of graphics card manufacturers (i.e. Do no evil). Do some search on eVGA and you'll be surprised to know how wrong your opinion was.

      ps. Sapphire is good, but their build quality and warranty is nowhere as good as eVGAs.

    13. Re:Man... by drxenos · · Score: 1

      My experience with ASUS was a GeForce card years ago. Half the time, installing drivers (or updates) would say there was no GeForce card in my machine. ASUS would never answer my repeated emails for support. Haven't bought anything from them since.

      --


      Anonymous Cowards suck.
  16. licensing? by Torstein+Haldorsen · · Score: 1

    Does this mean that AMDs will be the sole producer of ATI based graphics cards, or will ATI continue to licence their graphics tech to 3rd party chipmakers?

  17. Makes me uneasy by Mad+Merlin · · Score: 4, Interesting

    I can't see this being good for customers. As we all know, ATI's products tend to be miserably supported, though this hasn't been the case for AMD thus far. How will this affect the nForce line of chipsets? Given ATI's past I'd much rather have an nForce than whatever ATI kicks out.

    On the other hand, perhaps AMD will drag ATI out of it's rut, but I think it's just as probable that ATI will drag AMD down, and that's good for nobody.

    1. Re:Makes me uneasy by AtomicBomb · · Score: 1

      I am a bit more pessimistic. I don't think AMD needs to support much at this moment. CPU manufacturers will have to show you the instruction set. There are not much to hide. They don't have an Intel style chipset at this moment. In other words, they don't really need to write, support or open their device driver...

      By once ATI gets into the picture, the terrain changes. Arguably, ATI is bought by AMD not the other way round. But the expertise in supporting the chipset graphics card etc is in ATI. I believe it is more likely for the AMD to adopt the existing ATI support model... In other words, we can expect okay hardware with barely satisfactory software support..... I hope I am wrong.

    2. Re:Makes me uneasy by Anonymous Coward · · Score: 0

      "On the other hand, perhaps AMD will drag ATI out of it's rut, but I think it's just as probable that ATI will drag AMD down, and that's good for nobody."

      Whenever two companies with different product lines and cultures merge, it's difficult to say who will drag down whom --- but it certainly always is a risk. Given Intel's recent offensive, this one will be interesting to watch...

    3. Re:Makes me uneasy by quakeroatz · · Score: 3, Funny

      "ATI's products tend to be miserably supported"
      Oh tell me more, NVIDIA fanboy! Tell me a opposing tale of the wonderful NVIDIA happy land, with a gumdrop house on lollipop lane!

    4. Re:Makes me uneasy by Anonymous Coward · · Score: 0

      Every reasoned analysis I've seen says nVidia has better drivers. nVidia's linux drivers are most certainly superior. This is not to say one technology is better than the other, but from everything I've heard and seen, nVidia spends an awful lot more time on good drivers than ATI does.

      But then, I've refused to buy an ATI card ever since the quack3 debacle. I simply can't bring myself to trust ATI again. (For those of you who may not be aware, when Quake3 was used as the standard in benchmarking, ATI's drivers would detect whether quake3.exe was running, and if so, lower the graphics detail. Lower graphics detail == higher FPS. The idea was that most benchmarkers weren't looking closely at the screen while running timedemo's, so the benchmark graphs would show ATI as the better card and boost sales. People bringing home ATI cards would be running quake3 at low graphics. If you change the name of the executable to, say, Quack3.exe, you could then run the game at high detail. It was dirty and underhanded. Anyone going out and spending alot of cash on an ATI card would take it home and be playing the game with no better graphics than a three year old card. Why should anyone trust them ever again?)

      I'm saddened that AMD would buy a company who is willing to crap all over it's customers like ATI.

    5. Re:Makes me uneasy by Anonymous Coward · · Score: 0

      In case you couldn't tell, I was being sarcastic.

    6. Re:Makes me uneasy by zoomzit · · Score: 2
      Oh tell me more, NVIDIA fanboy! Tell me a opposing tale of the wonderful NVIDIA happy land, with a gumdrop house on lollipop lane!

      Homer: For your information, that was SARCASM.

      Marge: Well, Duh.

      I see your point about Nvidia, but I do have to agree with the basic premise of the parent that ATi's products are miserably supported (personally, I just did a clean install of Windows XP, downloaded the latest catalyst crap, and am now subjected to countless "graphic driver error" crashes.) One can only hope that AMD will rub off on ATi, and not the other way round.

      Personally, I don't see why everyone is excited about this merger. The market is consolidating and we have just lost a major player, this is always bad for the consumer. Plus, with AMD gobbling up ATi, AMD stands to lose access to Nvidia chipsets, which would effectively handicap AMD's ability to catch up in the processor race.

      I don't see any benefit to this merger except that it has driven the best GPU producer (Nvidia) to the current champ of processors (Intel). They might start making stellar product together, until they drive AMD/ATi out of business. From there, they will sodomize you, the consumer.

    7. Re:Makes me uneasy by moosesocks · · Score: 2, Informative

      No. The grandparent poster has a point.

      nVIDIA came out of nowhere about 5-6 years ago, whilst ATI has been firmly entrenched in the marketplace for a much longer time.

      nVIDIA was able to grow so quickly, because their products were faster, less buggy, and better supported than anything on the market at the time. ATI was just barely able to keep up, and everyone else bit the dust.

      The consumer-end graphics industry has been known for buggy drivers for almost its entire existance. nVIDIA's biggest innovation was not necessarily in hardware, but in software instead. By shipping a unified driver for all of its products, it was easy for them to fix bugs in the drivers years after the products they supported came out, and continually tweak those drivers. As a result, they wound up with a rock-stable platform that ATI took years to match.

      I'm not saying ATI's terrible (I'd consider them to be about average), but in terms of stability and product support, there's little doubt that nVIDIA leads the industry.

      AMD on the other hand, I see as a great company with great leadership. I hope that they can turn ATI around, and push out some innovative products. AMD's committed to open standards, and is known for its ethical business practices, not to mention that they tend to do much of their manufacturing and product development within the US.

      Who knows how this will all balance out?

      --
      -- If you try to fail and succeed, which have you done? - Uli's moose
    8. Re:Makes me uneasy by drinkypoo · · Score: 1
      (personally, I just did a clean install of Windows XP, downloaded the latest catalyst crap, and am now subjected to countless "graphic driver error" crashes.)

      I've since sold the machine but I had a machine with a 9700 XT. If I loaded catalyst control center, machine would bluescreen on every boot. Remove it, and the drivers work more or less properly. (As good as can be expected from ATI.) If you haven't already, remove CCC and see if anything improves.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    9. Re:Makes me uneasy by zoomzit · · Score: 1
      I've since sold the machine but I had a machine with a 9700 XT. If I loaded catalyst control center, machine would bluescreen on every boot. Remove it, and the drivers work more or less properly. (As good as can be expected from ATI.) If you haven't already, remove CCC and see if anything improves.

      Yeah, I actually removed CCC and reinstalled just the drivers yesterday, and stability has seemed to improve. It is painful to seem companies release software that causes their own hardware to crash...

    10. Re:Makes me uneasy by Vacuous · · Score: 1

      What you forget to mention is the more recent(But still a couple years old) debacle where NVIDIA released a forceware version that caused the card not to render scenes properly for an illigitimate FPS boost. The truth is, neither company is paticularly trustworthy, although I use ATI cards myself.

  18. AMD designs by bjb · · Score: 4, Insightful
    Interesting possibility:
    • Today: AMD has integrated memory controllers to get good memory performance.
    • Tomorrow: AMD has integrated video controllers to get good 3D performance.

    OK, so not very close to reality considering what would be involved. AMD bought into ATI because it wants to focus on CPUs, not chipsets.

    However, it does make for an interesting point of interest: the three primary components of PC architecture today are the CPU, GPU and chipset that bind the two together. AMD had two parts of the equation, and ATI has two parts as well, though one of these parts overlap. Now AMD is one company that has end-to-end solutions? There's got to be something interesting coming out of that marriage.

    --
    Never hit your grandmother with a shovel, for it leaves a bad impression on her mind...
    1. Re:AMD designs by Anonymous Coward · · Score: 0

      I upgrade my GPU every ~2 years and CPU every ~4. Unless merging them comes with a huge performance or price benefit I'm keeping them discrete.

    2. Re:AMD designs by PFI_Optix · · Score: 3, Insightful

      One thing that Intel has always done better than AMD is provide the "whole package".

      What I can buy from Intel:

      Server chassis + power supply
      Motherboard
      CPU(s)
      NIC
      RAID

      What I can buy from AMD:

      CPU(s)

      Small-medium OEMs are going to like Intel because it gives them one point of support for most of their major components. It also gives them a single "partner" with which to negotiate pricing; the larger volume of product means they can get overall better pricing.

      Taking on ATI might be AMDs move to start fixing that shortfall in their business model. If they put a solid OEM-friendly motherboard on the market, it will be a huge step in the right direction. With Conroe presently beating the pants off AMD's offerings, this is well-timed.

      --
      120 characters for a sig? That's bloody useless.
    3. Re:AMD designs by SchwarzeReiter · · Score: 1

      I dont think not selling motherboards and NICs is a shortfall of AMDs business model, because I dont think they can make much money with those activities. They just do it, because as a chipset makers they have to have a reference design for every chip they sell, and if they already have them, why not run the mill to produce some. They anyway have to support those reference designs, so it costs not a lot of additional money, and maybe they can sell enough to cover some of the design costs. But they cant make a lot of money in that businnes unit.

    4. Re:AMD designs by PFI_Optix · · Score: 1

      I dont think they can make much money with those activities. I doubt Intel makes a lot of money off their motherboards, but their line of NICs has become one of the best in the business and is likely quite profitable. But that's beside the point; they don't sell motherboards for the sake of selling motherboards, they sell motherboards because it helps them sell processors. I worked for a server OEM that had five major partners: Intel, Microsoft, Fujitsu, LSI, and Kingston. Microsoft: obvious Fujitsu: hard drives only LSI: RAID controllers (because they were the fastest by far of anything we tested) Kingston: memory, of course Intel: Everything else One thing that kept us from using AMD was the fact that we would have had to switch to using third-party motherboards AND adopt a new partner on CPUs. OEMs that can't make their own boards aren't crazy about having to use third-party boards for AMD chips because they tend to be inconsistent (for example, I recently worked on a motherboard where different batches had different sound chips, all under the same product name) and high-performance boards have excessive and expensive features that OEMs would prefer not to have. AMD would only benefit from marketing their own line of motherboards. Even if they broke even on the boards, they would almost be guaranteed to sell more CPUs.

      --
      120 characters for a sig? That's bloody useless.
    5. Re:AMD designs by evilviper · · Score: 1
      third-party boards for AMD chips because they tend to be inconsistent

      This is nonsense. If you buy dirt-cheap motherboards from PC-Chips, ECS, etc., you're going to be lucky to get something that works, let alone something that is consistent. You can't compare that to expensive Intel-branded motherboards.

      I made it a point to say "Intel-branded" because Intel is just the middle-man in this proposition. You can buy damn-near exactly the same motherboards directly from the companies that manufacture them for Intel (used-to be Asus, I haven't bothered to find who it is now).

      And similarly-priced AMD boards from the same high-end companies are every bit as consistent as their Intel-branded bretheren.
      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    6. Re:AMD designs by Anonymous Coward · · Score: 0

      From a procurement point of view it's much easier to deal with a single company, and you gain leverage.
      For example if you have an order for 1000 units, and they say they can only sell you 900 MB. In the case of Intel, they will not only lose out on the 100 MB, but also 100CPU, and whatever else you buy from them. They may do things like offer 100 higher value motherboards to make up the shortfall, because it will be less impact to them than missing out on the 100 CPU sales. With multiple vendors you don't have that kind of leverage.

    7. Re:AMD designs by evilviper · · Score: 1

      Yes, yes... I didn't argue with the "procurement" part of the post, only the "inconsistent" argument, which simply isn't true.

      And as for the latter, I expect you'll still come off cheaper with several different companies for AMD products, than for dealing with only Intel on more expensive items.

      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
  19. It WILL Be Good! by eldavojohn · · Score: 4, Insightful
    this could be real good if AMD's acquisition of ATI allows them to produce full chipsets in the same fashion Intel has with its Centrino line. let the competition begin!
    Yeah, the part that really sweetens the deal for us end consumers is that ATI will now get to benefit from the research that AMD inherits from IBM for chipsets. Hopefully ATI can make some better video cards with all the research that the other two have benefited off of. I hope that the same chipmaking technologies AMD has been using can now be used to improve ATI's GPUs and chipsets.

    Since (in my opinion) NVidia has taken the lead in GPUs, I hope that ATI will be boosted back into a competitive state and price wars ensue.

    Again, to me this is nothing but great news for the end-consumer.
    --
    My work here is dung.
    1. Re:It WILL Be Good! by Ubergrendle · · Score: 1

      Outside of the top end bleeding edge 7x00 cards -- you know, the ones that cost more than a motherboard + cpu + memory -- ATI competes fairly well on function/price. Its a good time to remember than probably ~80% of the market is made up by the low and mid tier.

      2 years ago it looked like Nvidia was dead meat, now they've come back strong. I only get worried when one of these companies can't get their sh*t together for 2 or 3 generations in a row...then you know they've stagnated.

      --
      John Maynard Keynes: "When the facts change, I change my mind. What do you do?"
    2. Re:It WILL Be Good! by tshontikidis · · Score: 1

      Many other too think that nVidia has the lead in the GPU market, if you are willing to pay.

    3. Re:It WILL Be Good! by Slack3r78 · · Score: 1

      Except we don't know how Nvidia will react to this news, especially so far as the chipset market goes.

      Fewer competitors is almost always a bad thing for consumers in the long haul, and we've just cut the number of players in the chipset market down to five (AMD, Intel, Nvidia, SiS, VIA), with two being the major CPU makers as well.

      The NForce chipset was what finally made the Athlon a viable platform for many, and if this move pushes Nvidia closer to Intel, it could be bad for AMD in the long run.

    4. Re:It WILL Be Good! by PhoenixPath · · Score: 1

      Meh...

      AMD chipsets? Oh yeah...I think I heard something about them ages ago... ;)

      This isn't eliminating competition, it's bolstering it. AMD was never really a market competitor in the chipset arena. AMD producing ATi Chipsets will improve (hopefully) ATi's offerings making them *more* competitive.

      Right now, if you want a decent AMD system, you're stuck with an (albeit good) nForce chipset. That's not quite what I would consider a wide array of options. Adding ATi/AMD to the mix with their own chipsets could *only* be a good thing..IMO.

    5. Re:It WILL Be Good! by dnoyeb · · Score: 1

      NVIDIA is already supporting every cpu intel releases so they are not weighted to AMD. Im sure this will help ATI accelerate their chipset production having AMD right there to help out. It will be bad if ATI gets too 'in-house'. I recall that NVIDIA kicked up the AMD chipsets about the time AMD decided to drop their own chipset. Lets hope that AMD does not get anti-competitive. intel does not seem to be anti-competitive toward NVIDIA and intel makes chipsets.

    6. Re:It WILL Be Good! by mabhatter654 · · Score: 1

      But the very nature of Intel integrated "everything" is horribly anti competitive to the chipset and graphics makers. Intel is bundling chipset, wireless, and graphics for only slightly more than the chipset.. and tying that to the OEM's marketing (centrino?) and processor costs. ATI is just about pushed out of the low end "default" graphics market for intel. I was just at the store and almost all the AMD notebooks and a good portion of the desktops were sporting ATI chipsets and graphics. That's where AMD is going with this. Now they can gain revenue by the addition of chipsets and graphics, not to mention sweeten the OEM deal by making the costs and engineering lower. I'd have liked to see Nvidia, but their prices are considerably higher than ATI's. Also, the OEMS seem to prefer dealing with ATI... many have been using ATI for a long time...that's another in road for AMD sales. I just hope AMD/ATI cleans up house and starts putting real graphics out there. This silly stuff of 3/4 of the new PCs on the shelf, not for gamming or multimedia (worse, even with MCE installed!) is getting awful. We could really dream (or smoke good stuff) and think AMD/ATI could almost start their own brand.. and actually make money selling a Linux "platform" but I'd doubt that.

    7. Re:It WILL Be Good! by Slaimus · · Score: 1

      Don't forget that nVidia chipsets are in many Opteron servers, as well as AMD "validated" business platforms. Now businesses will think twice before buying revenue cashcows like Opteron servers and Athlon64 desktops.

    8. Re:It WILL Be Good! by Slack3r78 · · Score: 1

      This is something I've brought up in conversation with some other people, actually. AMD started making a push on their "Corporate Stable Image Program" about this time last year. Certain AMD CPUs are labeled CSIP compliant. It's very clear that Nvidia has worked closely with AMD on this effort - I can buy NForce based CSIP boards all day long without having to go through too much effort.

      ATI is part of the program as well, but it always felt like they were playing more lip service to it than anything. I've yet to buy or even see an ATI based CSIP board. This compares to Nvidia and Nvidia partners like Asus who have been pushing the program pretty hard.

    9. Re:It WILL Be Good! by dfn_deux · · Score: 1
      "Anti-competitive" implies some sort of price fixing... If intel is able to cut their production costs by integrating everything and that allows them in the end to put it in consumer goods more cheaply then it isn't "anti-competetive" so much as it is just smart manufacturing....

      What's next, you gonna complain that GM manufacturing both engines AND transmissions is somehow problematic because it allows them to more cheaply platform engineer their drivetrains?!?!

      Good engineering/planning a monopoly does not make

      --
      -*The above statement is printed entirely on recycled electrons*-
    10. Re:It WILL Be Good! by mabhatter654 · · Score: 1

      but Intel is already under investigation for monopoly practices... if it wasn't for AMD's heroic attempt at competition, we wouldn't be as far as we are now. Intel is using their near monopoly on processors to force OEMS to purchase bundled chipsets, graphics chips and wireless cards. By it self, it's not wrong, but look how crappy the market is because for graphics because intel "encourages" OEMS to use their stuff instead of combining parts. Intel is nowhere near MS level of "evilness" but they still throw "discounts" to keep big customers in line from choosing too much AMD. Also, look how long it took Nvidia to get their "bus" license, and how quickly they come out and take ATIs away before the ink is even dry. Don't forget how intel got their graphics to start with, they "partnered" with Real3D to make add in cards.. .then hijacked the development team and "bought" the company out later.

    11. Re:It WILL Be Good! by obeythefist · · Score: 1

      Bah! IBM stole that research from SCO and integrated it unlawfully into the Linux kernel that SCO owns.

      --
      I am government man, come from the government. The government has sent me. -- G.I.R.
  20. Goodbye ATI? by XxtraLarGe · · Score: 5, Interesting
    "It's been a rumour for several weeks, but now it's confirmed: AMD buys ATI. What implications is this merger going to have for the hardware market?"

    I wonder if this means no more ATI cards in Macintosh computers, seeing as how Apple uses Intel now? Or, even more interesting, could it mean Apple switching over to AMD?

    --
    Taking guns away from the 99% gives the 1% 100% of the power.
    1. Re:Goodbye ATI? by Anonymous Coward · · Score: 0

      Well switching video cards seems a lot more straightforward than switching cpu/chipset for Apple. Going to be work either way, but hard to see how switching graphics vendors isn't going to be a magnitude easier.

    2. Re:Goodbye ATI? by Anonymous Coward · · Score: 0

      Out of the horses mouth (ie. ATI representative) they have declared that there will be continue support for Apple platform and definitely new products for the 'workstation' models in the future.

      Or as the rep put it, he's "thrilled" about the Apple migration to the Intel platform.

    3. Re:Goodbye ATI? by baldass_newbie · · Score: 1

      I was wondering the same thing. I can't see Apple running away from ATI just yet, but them moving to AMD is very unlikely as Intel has some integrated DRM stuff (Viiv) that Apple (AFAIK) is shooting for.

      --
      The opposite of progress is congress
    4. Re:Goodbye ATI? by WombatControl · · Score: 2, Insightful

      Why would AMD do that?

      No company would kill off a profitable product line just to spite their opposition. Undoubtedly ATI's deal with Apple is profitable, and just because Apple uses Intel processors doesn't mean that such a transaction is any less profitable than it was before.

      Companies don't act in that way, they look out for their bottom line. Unless there's something that would cause that business to become less profitable, ATI is unlikely to give up the block of sales they get from Apple. Is it better to cede that entire block of sales to the competition just because they don't use AMD processors? You don't win in business by reducing your sales, and having ATI graphics cards in Apples gives AMD/ATI a foothold in a very profitable market. It makes no business sense to give that up.

    5. Re:Goodbye ATI? by Johnny+Mozzarella · · Score: 2, Insightful

      Apple first began offering ATI graphics and then Nvidia and most recently intel graphics.

      iBooks always used only ATI graphics.
      iMacs have used both ATI and Nvidia graphics.
      PowerBooks have used both ATI and Nvidia graphics.
      PowerMacs have used both ATI and Nvidia graphics.

      The Mac mini and MacBook are currently using intel integrated graphics (high volume products)
      The MacBook Pro and iMac both currently use ATI graphics (high volume products)
      The PowerMac currently uses Nvidia graphics (low volume product)

      Apple has enjoyed the benefits of being able to pit ATI, Nvidia and intel against each other to get the best prices for their chips.
      I wouldn't be surprised to see AMD create a low cost chipset solution that we might see in a future Mac mini or iBook.
      Nothing is going to change in the next year but this will give AMD an opportunity to work with Apple and pitch it's wares.

    6. Re:Goodbye ATI? by EXMSFT · · Score: 1

      That and, as far as I know, AMD hasn't had their cup of the EFI Kool-Aid that Intel provided them. And until AMD supports EFI, no Mac OS on an AMD system.

    7. Re:Goodbye ATI? by mrxak · · Score: 1
      Nothing is going to change in the next year but this will give AMD an opportunity to work with Apple and pitch it's wares.
      That's a very good point. AMD would probably love to see Apple shipping AMD-based x86 systems. By buying one of Apple's longtime partners, this gives them some opportunity to exert some influence. That's not to say that we're suddenly going to see AMD Macs, but it does make it easier should Apple decide to go that route.
    8. Re:Goodbye ATI? by Speare · · Score: 1

      I think you misunderstand. AMD/ATI is probably *thrilled* to say that there's AMD in a MacBook Pro. But the fear is that Apple has likely gotten all sorts of agreement with Intel that they *not* include AMD components, and thus, Apple would dump AMD/ATI.

      Until now, Intel only had to worry about the processor market role that they played with Apple. Now they are sharing a motherboard with their competition, and the lawyers are likely to be sharpening their pencils and pondering how to get more money or more exclusivity from Apple.

      --
      [ .sig file not found ]
    9. Re:Goodbye ATI? by Anonymous Coward · · Score: 0

      Um, if WalMart couldn't get Apple to capitulate on iPod prices, I don't see how owning ATI will give AMD any leverage at all. Trying to twist Jobs arm with ATI leverage will likely result in AMD getting STEVED and they will lose all ATI sales to Apple for a few years.

    10. Re:Goodbye ATI? by six11 · · Score: 1
      I wonder if this means no more ATI cards in Macintosh computers, seeing as how Apple uses Intel now? Or, even more interesting, could it mean Apple switching over to AMD?

      I had the same thought. Once the acquisition settles down and A[MD|TI] can start using this to their benefit, I bet PC vendors will be able to get price breaks for going with both the CPU and the graphics chipsets/cards. If that would make Apple's hardware cheaper, that would be great for me. I'm po', you know?

    11. Re:Goodbye ATI? by caseih · · Score: 1

      Sure but until the Mactel towers are release with pci-express slots in them, there's no real support in OSX86 for nvidia cards at all. In fact there aren't even any drivers available yet, except from a third-party site. Currently OSX86 only ships with intel and ATI drivers. This says nothing about the future, though.

    12. Re:Goodbye ATI? by drinkypoo · · Score: 1
      Going to be work either way, but hard to see how switching graphics vendors isn't going to be a magnitude easier.

      I'm sure AMD would be more than willing to loan them a grip of people to help them get spun up, and Apple does eventually redo their motherboards (though apparently not necessarily with every "generation" - remember the early PowerPC upgrades? Some of those machines struggled along with the prior generation's processor bus.)

      This would be especially appealing to Apple if they could suddenly get a complete solution from one single vendor, guaranteed to all work together, etc etc.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    13. Re:Goodbye ATI? by mrxak · · Score: 1

      I meant "influence" in a friendly way. Like some AMD people talking to Apple about GPUs and just happen to bring up the subject of AMD's latest advancements in CPUs (perfectly plausible, since I'm sure ATI+AMD engineers will be sharing technology now). AMD now has a channel of communication open that they didn't have before. And they could always say, give a special deal on Radeon X2900s or something with every AMD-based system down the line.

    14. Re:Goodbye ATI? by Mistah+Blue · · Score: 1

      Actually senior executives routinely make stupid decisions to kill off products that make money. Remember, the fat cats only care about pumping up the stock price for their options and the quarterly results. These folks are generally not engineers and don't care about nor do they have a clue about making quality software or hardware. I'd love to be proved wrong here, but wouldn't hold my breath. Big acquistions like these are generally never good for the acquired company.

    15. Re:Goodbye ATI? by Anonymous Coward · · Score: 0

      Wouldn't EFI be supported by the motherboard chipset rather than the CPU? Assuming that is indeed the case and given that AMD doesn't yet make motherboard chipsets, how would AMD support EFI?

  21. Intel has killed gaming...but AMD has restored it! by ActiveMatx · · Score: 2, Interesting

    This will sorta relieve some of the high-stress factor from the "Intel has killed gaming" theory. In which, most business and "consumer" machines that come with fast intel processors but crappy integrated intel graphics is a joke. These users think "hey I got a Penitum 4, 3.2ghz, I am going to go play Half-Life 2, only to not meet the minimum requirements. With AMD releasing PC's combined with low-cost ATI chips imbedded into their "consumer grade" PC's this could have a strong uproar towards the PC gaming market. PC's will get cheaper, and more people will have a computer that is remotely possible of gaming (unlike in the past). I see this merger as a good thing. We know that ATI will continue to run neck and neck with Nvidia which is a good thing, because if ATI dropped out of the ball game, then Nvidia would hold the 3d graphics chip as a monopoly. (unless you consider Savage and Matrox competition)....

  22. The interesting thing to watch will be... by Glock27 · · Score: 5, Interesting
    NVIDIAs response. Will NVIDIA no longer support AMD processors to the same level? The shocking thing to me about this announcement is that nForce chipsets are the best chipsets for AMD64. Also, NVIDIA driver quality across the board is better than ATI.

    So, we'll see how this shakes out. If, as others have said, AMD forces ATI to produce better drivers, and good Linux drivers, that may be a good outcome...

    The other interesting aspect is (as it often is) Apple. Now AMD gets an instant slice of the Apple pie (sorry) since ATI makes most current Apple graphics chips. Interesting development there... Intel can't be happy.

    I suspect the tension level just notched up at NVIDIAs headquarters as well.

    --
    Galileo: "The Earth revolves around the Sun!"
    Score: -1 100% Flamebait
    1. Re:The interesting thing to watch will be... by spyrochaete · · Score: 1

      It's also interesting that the new AM2 architecture supports SLI out of the box, but Intel still has no dual core solution for SLI. I thought Nvidia and AMD were buddy buddy. Can AMD go both ways?

    2. Re:The interesting thing to watch will be... by powerlord · · Score: 1

      Considering that Intel sells graphics cards but also works with both ATI and Nvidia cards, I don't see why AMD couldn't support Nvidia, in addition to AMD's own internal card.

      --
      This space for rent. All reasonable inquiries will be entertained at proprietors discretion.
    3. Re:The interesting thing to watch will be... by spyrochaete · · Score: 1

      I don't really know if Intel is much of a competitor in the video card market. Intel graphics adaptors are for people who don't care about video performance. I don't even think you can upgrade from one Intel GPU to another. However, ATI and Nvidia are direct competitors and I'm afraid AMD might do something stupid to nurture their new acquisition - like screw up its currently excellent support for both Radeon and GeForce.

    4. Re:The interesting thing to watch will be... by Jungle+guy · · Score: 1

      Actually, I believe on nVidia headquarters there was some celebration. ATI, their direct oponent, no longer exists - it was bought by another company. AMD might be bigger than Nvidia, but they don't have the same focus and business strategy.

    5. Re:The interesting thing to watch will be... by MonsoonDawn · · Score: 1
      I don't really know if Intel is much of a competitor in the video card market. Intel graphics adaptors are for people who don't care about video performance.


      Huh!? Intel is the #1 supplier of desktop and laptop motherboards to OEM's. Nearly every desktop & laptop board Intel makes ships with integrated graphics. The vast majority of people who play games on their PC's do so with unmodified OEM boxes that use Intel graphics. Intel owns almost the entire market for graphics chipsets.
    6. Re:The interesting thing to watch will be... by Hawke666 · · Score: 1

      Exactly. The vast majority of people don't care about video performance.

    7. Re:The interesting thing to watch will be... by obeythefist · · Score: 1

      There will be many changes! But of course, Intel is again CPU king (and still outperforming AMD on price performance even after the AMD pricecuts). NVidia are making a nice NForce setup for C2D.

      We will know from the actions of Intel and NVidia if whatever AMD and ATI are off planning will be a decisive move or not. I trust that Intel and Nvidia corporate spies are going crazy right now.

      --
      I am government man, come from the government. The government has sent me. -- G.I.R.
  23. Nice deal, but there are soome who disagree... by nawcom · · Score: 1

    This is a great deal when it comes to combining AMD processing with ATI gpu technology. Yet almost every avid gamer and linux fan seems to prefer nvidia over ATI. nvidia seems to have better overall support. It's the only gpu brand name i can use to get accelerated graphics on my laptop running BSD. It took me a while to get over the 3dfx buyout; I was a big voodoo fan, but i then accepted nvidia as my gpu of choice. Don't give up nvidia! you still have your supporters... -nawcom

  24. Just to get it out of the way... by mushadv · · Score: 0, Offtopic

    blah blah blah Netcraft

  25. Worst case by botik32 · · Score: 2, Interesting

    Now that ATI is part of the AMD, the worst case is ATI division is given little attention, developers move to CPU core development, NVidia remains the only serious GPU vendor, and things go downfall from there.

    A second worst outcome is Intel enters a pact with NVidia, so next gen NVidia cards are so integrated with Intel chipsets that they do not run well on AMD. If you buy an AMD platform, you can only buy an ATI video card. If you buy an Intel platform, you are bound to NVidia. This would suck bad as well.

    I do not see good outcomes from this merger. Possibly better integration with AMD products is not something I as a customer would care about. ATI being focussed on GPUs only sounds much better for the customers than a division of a large computing devices company. At the very least, ATI is now somewhat concerned about Linux gamers, because it needs them. Who is to say that as part of a large and much more stable company, the ATI division will still give a rat's ass about Linux?

    The best outcome possible - With the backing of AMD's large patent portfolio, ATI division opens up the specs to all of their GPUs. Programming manuals. Device driver instructions. Linux drivers improve in leaps and bounds. Linux users switch to AMD+ATI. Windows gamers switch to Linux. Microsoft crumbles under its own weight and withers. Sightings of Belzebub trying on skates are reported.

  26. Ugggh by LaughingCoder · · Score: 5, Insightful

    I think the marketplace has been very well-served by the two dualities that existed before this move: ATI and NVidia beat each other's brains out, as did Intel and AMD. This new dynamic with 3 players does not seem, to me, to promise anywhere near as many benefits for us, the customers. Will ATI become more AMD-centric? Undoubtably. Will NVidia (which has been a great AMD booster) become less supportive of AMD processors? Probably. As this plays out, it seems to me that NVidia will basically be an Intel graphics house (including Macs), and ATI will melt into AMD, becoming mostly an internal chipset house. In the end we lose a very healthy competition between NVidia and ATI. We gain, perhaps, a stronger AMD to keep Intel honest.

    --
    The more you regulate a company, the worse its products become.
    1. Re:Ugggh by Chris+Burke · · Score: 1

      That only really applies to the chipset market, though. The standalone graphics cards are processor agnostic, being on one end of a standardized bus interface. I expect the GPU/graphics card battle to continue unabated for the forseeable future. Remember that the customer of ATI and NVidia are OEMs and retail purchasers who would be happy to pair an AMD processor with an NVidia card.

      Now as far as chipsets are concerned, it is probably true that NVidia will not support AMD as strongly as it has. However this is not a foregone conclusion. NVidia wants to sell chipsets, and if they make the best AMD chipset then OEMs will use it just like they'd use NVidia graphics cards in an AMD system. If NVidia refused to make a high-performance AMD chipset, then OEMs would just buy the ATI/AMD chipset or some other one. In other words, unless they thought that the OEM would stop using AMD so that they could use an NVidia chipset, then all they're doing is preventing themselves from making a sale, essentially giving up chipset marketshare.

      The key here being that AMD continue to support open standards like Hyper Transport, and not erect artificial barriers like Intel -- e.g. locking out chipset competitors with IP as in the case of Via and the P4 bus, or requiring OEMs to purchase Intel chipsets along with CPUs as in the case of Centrino. As long as that holds true, then the OEMs are going to pick what they consider the best chipset for their product, and NVidia thus has an incentive to grab as much of that marketshare as they can.

      --

      The enemies of Democracy are
  27. Chipsets by ArcherB · · Score: 1

    Well don't expect the same level of cooperation between AMD and nVidia that we've seen these past few years. For some reason I don't see nVidia getting terribly excited about making chipsets for their number one competitor.

    --
    There is no "I disagree" mod for a reason. Flamebait, Troll, and Overrated are not substitutes.
  28. I wonder how this will play out... by Xest · · Score: 2, Insightful

    ...for people like me who were in the AMD/nVidia fanclub? I've always had countless problems with ATI cards both at home and work, generally down to driver issues so I really don't want to switch to ATI, I'd personally rather go the Intel/nVidia route if this will have some adverse effect on using nVidia kit with AMD kit. I'm not sure this is good for the market either if there is some kind of lock in to ATI if you used Intel, it was kind of nice knowing you could choose between 2 processor manufacturers and 2 graphics chipset manufacturers, now it kinda feels like the choice has been dented somewhat in that you can't mix and match so well.

  29. Wishful thinking by Alioth · · Score: 1, Interesting

    The wishful thinking is that now ATi are owned by AMD, they might produce 3D hardware which they publish the hardware interface so we can ave open source graphics drivers. But I'm sure it'll never happen.

  30. And the cycle begins again. by Criffer · · Score: 2, Interesting

    The Wheel of time has turned again. GPUs are now general-purpose massively-parallel computers; they will be folded back into the CPU core, so that the general purpose CPU gains massive parallelism. Kind of like SIMD, but on the order of a million operations per instruction instead of 8.

    The next 10 years will consist of a new type of external graphics hardware being built, which will of course, be folded into the CPU at the end of it.

    1. Re:And the cycle begins again. by Lord+Kano · · Score: 1

      The next 10 years will consist of a new type of external graphics hardware being built, which will of course, be folded into the CPU at the end of it.

      I'm thinking that the next big thing to go through the cycle will be Physics Processing Units. That is if they pick any momentum in the marketplace at all.

      LK

      --
      "Hi. This is my friend, Jack Shit, and you don't know him." - Lord Kano
    2. Re:And the cycle begins again. by Criffer · · Score: 1

      I was going to mention physics processing units, but I really don't think they will take off. Not in the way graphics cards have. Physics just isn't as sexy.

      That said, I think that some sort of physical calculation support will end up in the CPU. Given that the GPU is basically a massively parallel computer, and that people have already done physics simulations on GPUs, I predict that the GPU on CPU technology will be used for physics more than any external physics card. It's not quite the same as folding the technology from an external device into the CPU, it's a case of reusing existing CPU grunt for a new purpose.

  31. See Hackers Dictionary: "Wheel of reincarnation" by kriegsman · · Score: 5, Informative
    AMD needs GPU functionality on the CPU.

    See the entry in the Hacker's Dictionary / Jargon File for "Wheel of reincarnation":
    wheel of reincarnation: [1968] Term used to refer to a well-known effect whereby function in a computing system family is migrated out to special-purpose peripheral hardware for speed, then the peripheral evolves toward more computing power as it does its job, then somebody notices that it is inefficient to support two asymmetrical processors in the architecture and folds the function back into the main CPU, at which point the cycle begins again.

    Several iterations of this cycle have been observed in graphics-processor design, and at least one or two in communications and floating-point processors. [...]


    -Mark
  32. Comment removed by account_deleted · · Score: 1

    Comment removed based on user account deletion

  33. Re:Intel has killed gaming...but AMD has restored by Andrew+Kismet · · Score: 1

    What kinda nonsense theory is that? My Intel Core Duo LOVES gaming, and the nVidia card that came with it likes HDR lighting, too.

  34. Shame it's not a merger by Rik+Sweeney · · Score: 0

    then they could have called themselves

    ATMID

    or

    DAMIT

    Sorry, that's the best I could come up with...

    1. Re:Shame it's not a merger by Eliman · · Score: 1

      ADMIT

    2. Re:Shame it's not a merger by The+Ribena+Kid · · Score: 1

      They could then have had the marketing slogan "ADMIT IT, it's better!"

  35. Intel's baby gfx chips suck by Anonymous Coward · · Score: 0

    So maybe if they did buy NVidia, they'd go to fast-but-unreliable instead?

  36. awesome by syrinx · · Score: 0, Redundant

    I've used only AMD CPUs and ATI graphics cards for 5-6 years now. This can only be good news.

    --
    Quidquid latine dictum sit, altum sonatur.
  37. Stock by Miros · · Score: 1

    So, AMD's stock, up or down in the short-term? They've been taking a bit of a beating over the past 3 months.

    1. Re:Stock by Anonymous Coward · · Score: 0

      OMG! I got the ticker wrong! I'm long Altai Resources!!

    2. Re:Stock by tpjunkie · · Score: 1

      AMD: Down nearly 10% already today
      ATI: Up over 20% already today

    3. Re:Stock by limabone · · Score: 1

      It all depends on whether the analysts think they are overpaying for ATI...down if they do, up if they don't. Long term though this seems like a good idea for AMD.

      I have an 3800X2 and an X1900XTX in my box and I just caught them making kissy kissy faces at each other!

  38. Towards GPL drivers for ATI cards ? by jeremie_z_ · · Score: 1

    Let's hope that AMD will make the smart move of opening up the specs of its ATI video cards, thus encouraging the development of GPL drivers. I am convinced that this could be a decisive step in the "war" with Nvidia...

    1. Re:Towards GPL drivers for ATI cards ? by TheRaven64 · · Score: 1

      Intel has drivers for everything up to the i915 (the i950 series are a work in progress) included in the DRI project, which has an MIT-style license. I would take MIT-licensed drivers over GPL any day.

      --
      I am TheRaven on Soylent News
  39. What about macs by tezbobobo · · Score: 1

    What I'm really concerned about is whether we'll still be getting ATI graphics in Apples. I've always had better performance from the ATI cards in Apples. This could limit Apples to NVidia - bad if they fall behind.

    1. Re:What about macs by Yvan256 · · Score: 1

      Apple needs the GPUs that have the best/fastest OpenGL. Isn't that nVidia? Isn't ATI more Direct X oriented?

    2. Re:What about macs by mrxak · · Score: 1

      It's been my experience with the few generations I've gone through, that ATI makes a cheaper/faster graphics card for the Mac high-end. They tend to last longer too. But anyway, don't worry about it. AMD/ATI has no desire to give up their profits selling GPUs to Apple. They'll keep fighting just as hard as nVidia, and Apple will continue to play all three (I'm including Intel's integrated chips in the mix) against each other for the best performance and cheapest price.

  40. Don't believe it myself by 0123456 · · Score: 4, Interesting

    Seems highly unlikely to me that they'd stick a GPU into the CPU. Modern GPUs are a similar size to CPUs (if not larger) and need much higher memory bandwidth... so you'd be doubling the size of your CPU and you'd need a 256-bit 1GHz+ memory interface. And then the 'high end' users would just go and buy a PCI-Express card when the next generation came out, making the whole thing a total waste.

    I could see perhaps that they'd stick a cheap and crappy GPU into a cheap and crappy CPU for the low end of the market, but with Vista coming out with all its eye-candy that may not even be viable for rendering the Vista desktop, let alone games.

    1. Re:Don't believe it myself by Whibla · · Score: 1

      "Modern GPUs are a similar size to CPUs (if not larger)"

      Absolutely true, there are more transistors on your top-of-the range graphics card, than there are on most processors...

      "you'd need a 256-bit 1GHz+ memory interface"

      Hello? AMD chips have an integrated memory controller, with the AM2s having support for DDR2 memory, running at 667 MHz. So unbuffered memory, running in dual channel mode runs at a speed of approximately 10.7GB/sec. In comparison current graphics technology runs at about 6.4GB/sec.

      Either one is fast enough (ouch - take away my \. membership), but personally, when I'm building my next pc, I'll go for the faster one given the option.

    2. Re:Don't believe it myself by Pulzar · · Score: 1

      Seems highly unlikely to me that they'd stick a GPU into the CPU. Modern GPUs are a similar size to CPUs (if not larger) and need much higher memory bandwidth... so you'd be doubling the size of your CPU and you'd need a 256-bit 1GHz+ memory interface. And then the 'high end' users would just go and buy a PCI-Express card when the next generation came out, making the whole thing a total waste.

      GPU/CPU combination wouldn't be meant for high-end graphics systems, but the low-end ones. It's just the next step after integrating graphics into the chipset, which has already proven to be a huge (financial) success. Low-end graphics are already sharing the system memory bandwidth with the CPU, in both integrated graphics and low-end discrete cards (TurboCache, and whatever ATI's similar solution is called)...

      --
      Never underestimate the bandwidth of a 747 filled with CD-ROMs.
    3. Re:Don't believe it myself by 0123456 · · Score: 3, Informative

      "Hello? AMD chips have an integrated memory controller, with the AM2s having support for DDR2 memory, running at 667 MHz. So unbuffered memory, running in dual channel mode runs at a speed of approximately 10.7GB/sec. In comparison current graphics technology runs at about 6.4GB/sec."

      Hello? My 7800GS card has a memory bandwidth of 40GB/second from on-board RAM. It would be utterly crippled by a measly 10GB/second shared with the CPU.

    4. Re:Don't believe it myself by Joe+The+Dragon · · Score: 1

      HTX based ATI video cards are a better idea then one in the cpu

    5. Re:Don't believe it myself by Lesrahpem · · Score: 1

      and you'd need a 256-bit 1GHz+ memory interface

      I take it you haven't looked at AMD's Opteron chips much, have you?

    6. Re:Don't believe it myself by Sloppy · · Score: 1
      Modern GPUs are a similar size to CPUs (if not larger) and need much higher memory bandwidth...

      Modern CPUs need way more memory bandwidth than they're getting. Go ahead and increase the combo CPU/GPU chip's access to memory in order to make it sufficient for graphics -- I'm sure the CPU won't mind.

      --
      As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
    7. Re:Don't believe it myself by i7dude · · Score: 2, Insightful

      ignoring the obvious chipset arguement that seems to arise. could another posibility be that amd wishes to improve its floating point design? i mean, gpu's are essentially just huge application specific floating point units. i dont really have enough information or understanding of the extent of amd's current designs to know, but it would seem they could gain a lot of ip in that area.

      dude.

    8. Re:Don't believe it myself by Pulzar · · Score: 1

      HTX based ATI video cards are a better idea then one in the cpu

      Well, that's pretty much what a Radeon Xpress chipset for AMD is, if you don't plug in an external card into its PCIE slot -- an HT video card. You still have all that latency up and down HT to deal with, though, vs. a solution in the CPU itself.

      --
      Never underestimate the bandwidth of a 747 filled with CD-ROMs.
  41. I'm looking at my AMD/nVidia dev machine and my Intel/ATI laptop and I'm thinking... oh crap.

    1. Re:Uh oh by Prof.Phreak · · Score: 1

      I'm looking at my AMD/nVidia dev machine and my Intel/ATI laptop and I'm thinking... oh crap.

      Exactly what I was thinking. This is a pretty weird combination of corps. I would've thought AMD would sooner buy nvidia than ATI.

      --

      "If anything can go wrong, it will." - Murphy

    2. Re:Uh oh by OverflowingBitBucket · · Score: 1

      Exactly what I was thinking. This is a pretty weird combination of corps. I would've thought AMD would sooner buy nvidia than ATI.

      I can understand the move now it has been made but I never would have guessed it. The dice fell badly for my particular combo. My dev machine has AMD for the value and nVidia for the Linux drivers. My laptop has Intel for the availability and ATI for the price. Whoops!

  42. NVIDIA on MacBooks now? by tbcpp · · Score: 0, Offtopic

    Does this mean we actually might get a Nvidia powered MacBookPro? If so I actually may buy one.

    --
    Man is the lowest-cost, 150-pound, nonlinear, all-purpose computer system which can be mass-produced by unskilled labor.
  43. it's simple, really by neersign · · Score: 3, Insightful

    I read thru most of the comments on this page, and several people came close to what I think the real reason for this deal is, but no one nailed it. To me, this is a simple example of business 101. AMD has always been a niche vendor. Recently they have begun to spread out, but it is obvious from all the comments on this page that they are still a "gamers" chip. Where Intel and Dell made it big was low-end, mass sale business computers. Intel has their crappy but good enough integrated video chipset which is a part of the vast majority of motherboards. In order for AMD to really be a big player, it needed to a) build it's own integrated chipset from scratch or b) buy a company that already makes integrated video chipsets. Option b won, and while it might cost more initially, it should pay off in the long term.

    I believe this will not stop nVidia from making nForce boards, and it would be stupid of AMD to stop production of ATI 3d cards. I think this may increase the quality of ATI's support for Linux, but I don't think it will be anything drastic.

    1. Re:it's simple, really by EmagGeek · · Score: 3, Insightful

      "AMD has always been a niche vendor."

      Are you smoking crack? AMD has most certainly NEVER been a niche vender...

      CPUs
      FLASH
      SRAM
      PLDs
      Embedded Processors
      Microcontrollers
      Ethernet Controllers and PHYs

      What niche exactly are you talking about here?

    2. Re:it's simple, really by SenseiLeNoir · · Score: 1

      Well said. I have an old Compaq EVO NV1015 Laptop. I dont know if you rememebr, but this particular laptop was a test case by Compaq/HP/AMD/ATI. The idea was to build a laptop that had reasonable power, yet cheap enough for bulk business purchases. The project was done by the four companies working together on a reference platform.

      The result: An AMD Mobile Athlon XP 2000+ with the ATI Radeon Mobility chipset, coupled with 256MB RAM, a decent keyboard, and sound system, in a very smart package consistig of Windows XP pro, and the bare minimum of applications (no bloat), all sellign for the price of £600 inc tax and delivery in April 2003, which is actually a VERY good deal at the time.

      I added a further 256MB of ram to this. And have used it as a development machine, server, Linux dual boot, and general usage.

      This laptop is STILL in use 3 and a half years later, runs Ubunto flawlessly, and as for the Windows side, i STILL have not reformatted/reinstalled it. Recently I Swapped the Hard Disk, but rather than reinstalling windows, i simply used Ghost to clone it. The only issues with that laptop is one of the fans are slightly noisy (easily replaceable), and the LiIon battery doesnt last as long as it used to. But it still has a VERY usable performance (more than enough for me), and for its price, has very good value.

      This merger will hopefully bring more of these to the market.

      --
      Have a nice day!
    3. Re:it's simple, really by wild_berry · · Score: 1

      If you credit Charlie at the Inquirer with anything, the Sun Niagaras imply a future where lots of simple processing units in the CPU will replace the large-scale OoOE at present, which would also be an easy place to put unified or vertex/pixel shader support too. He claims Intel has been working on it for a while already, which clarifies why those comments were made about Intel stifling gaming by the head of Epic Megagames as actaully being about Intel delaying things so as to allow their alternative solution to leave their test labs.

      Charlie's convinced (that there is a conspiracy to swap all of the world's bits from 0 to 1 and 1 to 0 at midnight on the 31st September, but that's a side issue, and) that AMD/ATi will leave the chipset business to nVidia so as not to anger them. I'm not sure of that one.

    4. Re:it's simple, really by Mad+Marlin · · Score: 1

      Compared to Intel, AMD is a niche vendor . Intel's revenue last year was $38.83B, over six times the $5.85B revenue of AMD. Intel makes everything AMD does, everything in your list, and a whole lot more too.

    5. Re:it's simple, really by EmagGeek · · Score: 1

      The fact that Intel sells a broader scope of product does not logically make AMD a niche vendor. By your logic, a company that sells 1E6 different products is a niche vendor if there exists in the market another vendor that sells 1E6+1 products.

      Why don't you try to make an argument that postulates that AMD is focused on a niche market, and then provide some evidence that supports it? Oh, I know, because that's not possible. AMD has too broad a product offering to be considered a niche vendor. If there were only a CPU company, fine. Only an Ethernet PHY company, fine. But, they sell dozens of product lines in various different markets. Thus, they cannot be considered a niche vendor.

    6. Re:it's simple, really by Mad+Marlin · · Score: 1

      Compared to Intel, their primary competitor , they are. If you run a fruit stand across the street from a full supermarket, then you are a niche vendor .

  44. Nah, weren't you reading yesterday? ... by powerlord · · Score: 3, Interesting

    The linux crowd (or at least a vocal minority of them), don't want drivers, they just want documentation for the card, they'll make their own drivers.

    On the other hand, releasing either open source drivers, or a combination of binary drivers, along with documentation (so those who want to write their own CAN), would certainly be the best of both worlds.

    --
    This space for rent. All reasonable inquiries will be entertained at proprietors discretion.
    1. Re:Nah, weren't you reading yesterday? ... by powerlord · · Score: 1

      An added bonus I just thought of.

      Release Open Source drivers, and even if there is something Nvidia can use, they theoretically can't without opening up THEIR drivers.

      (Yes, they could theoretically just try to steal code, but I am not sure how likely that is to happen, or how successful that might be)

      Possible End result? Open Source drivers for ATI AND Nvidia. ... Of course, this is assuming there is something in the drivers that is worth it to both of them ... or whichever one doesn't open source first.

      --
      This space for rent. All reasonable inquiries will be entertained at proprietors discretion.
    2. Re:Nah, weren't you reading yesterday? ... by Anonymous Coward · · Score: 0

      As a Linux user, I want drivers. If ATI would release open-source ones, that would be nice, but if they can't do that without giving away proprietary workings, at least make them binary. What I care about is whether I can develop and run my apps on Linux.

    3. Re:Nah, weren't you reading yesterday? ... by LWATCDR · · Score: 1

      "The linux crowd (or at least a vocal minority of them), don't want drivers, they just want documentation for the card, they'll make their own drivers."
      Not the Linux users want drivers that work and are easy to install.
      I for one am ticked at both the Linux Kernal developers and at Nvidia.
      I am ticked at the Linux kernal developers for NOT having a stable binary driver interface. And yes I have heard all the reasons but they all come down to "We want to force the hardware vendors to release open source drivers". What they will not say is that the lack of a binary interface prevents hardware vendors from packaging binary drivers in the box or on the Internet that are easy to install and use. Even if they release they RELEASE THEM AS OSS.
      I am ticked at nVidia because their binary drivers don't work. They will not load if you are using Xen.
      While I freely admit that I do like nVidia video cards and AMD cups that Intel had released the specs to their video chips. While they might not play the latest and greatest video games for a Linux desktop an Intel video card is the pay free way to to.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    4. Re:Nah, weren't you reading yesterday? ... by XMilkProject · · Score: 1

      And I don't know how many times ATI has to tell the damn linux crowd that they can't release the source code to their drivers because they are using various proprietary technologies under contract. If they release this information they would be breaking the law.

      It has been recommended that ATI release the information that it can, and the various proprietary technologies would not be utilized, but then ATI's cards would perform badly with these open source drivers, and they can't support that either.

      --
      Big ones, small ones, some as big as yer 'ead!
      Give 'em a twist, a flick o' the wrist...
    5. Re:Nah, weren't you reading yesterday? ... by mjrauhal · · Score: 1

      Not the Linux users want drivers that work and are easy to install.

      Disregarding your faulty English, FOSS mainline drivers would indeed provide this, being pretty much integrated into the kernel and X.org.

    6. Re:Nah, weren't you reading yesterday? ... by Brandybuck · · Score: 1

      Then why don't the release specs to the stuff they aren't hostage to? We don't need the source code, just the specifications. Geez.

      And next time, maybe they won't be so stupid as to sign an NDA on a freaking *specification*!

      --
      Don't blame me, I didn't vote for either of them!
    7. Re:Nah, weren't you reading yesterday? ... by LWATCDR · · Score: 1

      How would FOSS driver provide drivers that work?
      Do you actually think that every FOSS driver and piece of software is bug free, feature rich, and easy to install?
      And even if they where the versions of OSS software available in each distro's often lag the version of the software available from the authors. The lack of stable binary interface still would mean that a new version of an nVidia driver would REQUIRE the end user to compile it. Which may require the end user to have the correct version of gcc installed and the correct version of the header files...
      So not FOSS drivers do not guarantee a working easy to install and update driver!
      So much for the the if they make if FOSS then every thing will be fine line of thought.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    8. Re:Nah, weren't you reading yesterday? ... by mrchaotica · · Score: 1
      Do you actually think that every FOSS driver and piece of software is bug free, feature rich, and easy to install?

      Easy to install, yes -- checking an option in menuconfig is a hell of a lot easier than downloading and installing a binary blob separately!

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    9. Re:Nah, weren't you reading yesterday? ... by mrchaotica · · Score: 1
      ...but then ATI's cards would perform badly with these open source drivers, and they can't support that either.

      Who cares?! Mediocre open-source support is better than no open-source support at all, and it would at least provide a base for the community to improve upon. In my opinion, nVidia and ATi really have no excuse!

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    10. Re:Nah, weren't you reading yesterday? ... by LWATCDR · · Score: 1

      But not as easy to install as downloading a binary deb or RPM.
      And how is it easier? If a new version of the driver is released you still have to download the source tarball, untar it in the right location then run menuconfig so no not really.

      We do have binary drivers. The system in use so far has not prevented them.
      A binary driver interface would make it easier for the end user to install all drivers.
      Here is yet another case where the ease of use for the user takes a back seat to the interests of the developers. Kind of like DRM.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    11. Re:Nah, weren't you reading yesterday? ... by mrchaotica · · Score: 1
      But not as easy to install as downloading a binary deb or RPM.

      True, but irrelevant. If you're using a distro, having a Free Software driver would allow the distro maintainers to build it into the kernel to begin with. There wouldn't be a deb or RPM to download, because the hardware would have been autodetected and the accelerated driver module loaded at startup.

      Remember, this isn't like Windows: the way it's supposed to work is that the user should never install drivers at all, because they're already there. You've really got your perspective backwards, because binary blobs make it harder for users.

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    12. Re:Nah, weren't you reading yesterday? ... by LWATCDR · · Score: 1

      No you live in a fantasy land.
      All software gets updated. Do you think your distro has the very latest kernal patches, device drivers, and software versions in there repository?
      The average distro has thousands of packages. They can not keep them as up to date as the actual authors can.
      A good example I just ran into was Blender.
      It is in the OpenSuse repository but it is two minor revisions behind the current version. So I want to have the current version I have to download it from Blender.org and install or compile it from source. That isn't a problem for me. For the average user it is a problem.
      Another example would be any hardware vendor that wants to release a piece of hardware for Linux. When I set up the first Linux server at the office we got a 3Ware raid controller because it worked with Linux and had an open source driver. Well when we upgraded suse the driver stopped working. Because of the Kernel update the old driver wouldn't load. We had to recompile the driver for the new Kernel. Again not a big deal for me but for a lot of users it would have been a nightmare. Where is a great example of where a stable binary driver interface would have solved the problem. The module would have just loaded and I would have never need to recompile it. The hardware company I bought from was trying to support FOSS software, they had released the specs, they wrote a driver and released the source. The Kernel developers hadn't gotten around to putting it into the kernel at that time. So we had a user doing everything perfectly, a company doing everything prefectly, and still none of the oh it is in the kernal already because it is FOSS.

      No company is going to want to launch a new product for Linux if they have to ,give the kernel developers the specs and wait for them to feel like writing the driver! They are also not going to want to submit patches to the kernel and wait for them to make in into the distributions. Or depend on the distributions to distribute bug fixes to their device drivers when they feel like it.

      The ideal solution is one where the vendors release the specs and the drivers as FOSS and can release binary updates that just work for their users.
      Of course it would be a perfect world if every patch and fix to a device driver was instantly looked at and added to the kernel source tree. And then if every distro instantly updated all their users to that new source tree. But that isn't going to happen ever.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    13. Re:Nah, weren't you reading yesterday? ... by mrchaotica · · Score: 1
      For the average user it is a problem.

      No it isn't, because the "average user" doesn't care if he's using the absolute latest minor revision of every program. He's just going to keep using the version his distro provides (even if it's two minor revisions behind), blissfully unaware that the new version is even "available" (since it isn't available for him).

      When I set up the first Linux server at the office we got a 3Ware raid controller because it worked with Linux and had an open source driver. Well when we upgraded suse the driver stopped working. Because of the Kernel update the old driver wouldn't load.

      Again, that's irrelevant because an "average user" would not use drivers that weren't part of the distro-provided kernel to begin with. As far as the average user is concerned, if it's not in the distro's kernel, it's not supported.

      No company is going to want to launch a new product for Linux

      Who ever said they wanted companies to make stuff "for Linux"?! I certainly don't expect them to do that, nor do I expect support. All I ask is that the company provide detailed specs for the hardware I buy (which I would like to have regardless of the OS I use). Besides, if you consider the number of different OSs out there, it's unreasonable to expect the vendor to make drivers for any OS; at most, I would hope for just an open-source reference implementation to refer to.

      The ideal solution is one where the vendors release the specs and the drivers as FOSS and can release binary updates that just work for their users. Of course it would be a perfect world if every patch and fix to a device driver was instantly looked at and added to the kernel source tree. And then if every distro instantly updated all their users to that new source tree. But that isn't going to happen ever.

      What are you trying to argue here? It sounds like you're saying that because there might be issues with compatibility between versions that it automatically means the community is completely incapable of providing any usable drivers at all and that we should not even bother to want them. This is called a "false dichotomy," and is a fallacy.

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

  45. Just one question by martinultima · · Score: 2, Insightful

    From what I've heard, AMD tends to be pretty Linux-friendly, and very helpful to open-source developers who want to, say, implement AMD64 support and that kind of thing – so will this mean that ATI might start giving a damn about us too? I dunno, probably way too far-fetched, although I can't stand how my brand-new Athlon 64 box can't run 3D because ATI's stupid drivers pretty much don't work on my distribution... either way, though, so long as at least one of them keeps churning out good chips, more power to 'em!

    --
    Creative misinterpretation is your friend.
  46. Re:Intel has killed gaming...but AMD has restored by MrSquirrel · · Score: 1

    You do know that AMD doesn't make PC's, right? Just because big-name PC manufacturers (Dell, HP, etc) use Intel chips in their machines with onboard graphics, it doesn't make it Intel's fault that the machines can't game. If the manufacturers were using AMD chips, they'd still have super-lame-o-vision onboard graphics. Manufacturers care jack-diddly about performance, they just want specs with high numbers. ATI + AMD won't do anything to make PC manufacturers go "Hey, let's make machines that actually perform well instead of just having 'ohmygosh specs'".

    Anyone who wants to do high-graphic game can use their BRAIN and see that onboard graphics won't cut it... and there is a simple solution for them (if they don't want to build their own computer which is the smartest/cheapest way to go) -- it's called a graphics card; you can pick up a fairly decent one for $150, but if you're on a budget you can get ones cheaper that will still fit your need.

    --
    A computer once beat me at chess, but it was no match for me at kick boxing.
  47. AMD over-spending - good for Intel by hutchike · · Score: 2, Interesting

    Before this deal, AMD had about $3bn in cash. The Register says that Morgan Stanley will loan AMD about $2.5bn to get the deal done. This will leave AMD with no spare cash to fight the long war against the onslaught of Intel Core processors and upcoming quad-core Xeons - due this year. Disclosure - I just picked up 600 Intel shares at $17.50.

    --
    Zen tips: Pay attention. Don't take it personally. Believe nothing.
    1. Re:AMD over-spending - good for Intel by thegnu · · Score: 1

      Disclosure - I just picked up 600 Intel shares at $17.50.

      I actually picked up 600 shares of dust-bunnies over the weekend for free. The rich, of course, are alway emulating the poor.

      --
      Please stop stalking me, bro.
    2. Re:AMD over-spending - good for Intel by contiguously · · Score: 1

      Yes, they're paying about 4.5B to ATI... but then they get ATI, which means they get the assets they put into the acquisition back; not to mention ATI's assets.

      Good luck with your intel stock. There's a reason why it's at 17.50.

    3. Re:AMD over-spending - good for Intel by Tweekster · · Score: 1

      Well a bunch of people in ATI just lost their jobs and that will basically cover a good portion of the buyout so it is not like AMD is suddenly at zero dollars.

      --
      The phrase "more better" is acceptable English. suck it grammar Nazis
    4. Re:AMD over-spending - good for Intel by hutchike · · Score: 1

      You're right - no need to disclose, but I thought you might as well know about my unconscious bias in the AMD matter. I mean, were I feeling more balanced, I might have pointed out the future of the GPU and The Register's excellent perspective on it here. Anyhow, you're right - it's a bit "dickish" to disclose on Slashdot.

      --
      Zen tips: Pay attention. Don't take it personally. Believe nothing.
  48. AMD buying ATI by Anonymous Coward · · Score: 0

    Oh, great. Now it will be even more difficult to find a nice laptop with an AMD CPU and NVidia graphics chipset.

  49. Dupe? ;-) by Ajaxamander · · Score: 1

    http://slashdot.org/article.pl?sid=02/04/01/163621 3

    Thank you, thank you, I'll be here all week :)

  50. Anyone ever think... by Darundal · · Score: 1

    ...that maybe this is a multimedia push? Maybe this has more to do with their tuners, their All-In-Wonders then with their R520's and R480's. Maybe this is a a response to VIIV, only not half-assed.

  51. AMD not going to play nice with nVidia anymore?? by StarvingSE · · Score: 1

    I have to admit, I am an nVidia fanboy, and have used their chipsets and graphics cards for a long time now. I went with the AMD athlon xp/nforce2 chipset because of the stability it provided, and it has been so rock solid that I haven't felt to need upgrade the processor/motherboard for nearly 4 years now (need some speed, up that clock a little bit!)

    I wonder if this means that the combo made in heaven, amd processors with nforce chipsets, is going to be going bye bye thanks to this buyout?

    --
    I got nothin'
  52. Execs overly optimistic by rfunches · · Score: 4, Interesting

    AMD is covering the remaining $2.5b of the deal with a commitment letter from Morgan Stanley Senior Funding, with the debt secured by "a pledge of the capital stock of certain material units of the company, accounts receivable and proceeds from any sale by Advanced Micro of its equity interest in Spansion Inc." The CFO is overly optimistic that the company can get rid of that debt "quickly," without layoffs, and with savings of $75m and $125m over the next two years. DJ Newswires says ATYT will no longer work with Intel, and the execs say that they can make up the lost sales with the severing of Intel-ATI ties. Pretty lofty goals, I'd say.

  53. Re:Intel has killed gaming...but AMD has restored by ActiveMatx · · Score: 1

    I am just the messenger here. I am only speaking on behalf of a lot of inside journalist and key-note speakers. Here are a few different sites speaking about the "intel has killed gaming" idea. 1. http://www.joystiq.com/2006/07/12/epics-mark-rein- intel-is-killing-pc-gaming/ 2. http://blogs.guardian.co.uk/technology/archives/20 06/07/12/is_intel_killing_pc_gaming.html 3. http://arstechnica.com/news.ars/post/20060712-7247 .html

  54. NForce by Anonymous Coward · · Score: 0

    What does this mean for us that have the NForce hardware and AMD chipsets? Will NVidia still build and release drivers for this platform or start to phase out this technology because of the ATI/AMD merger? I have a feeling I am stuck with orphaned hardware because of this.

  55. Here's one obvious thing: by MtViewGuy · · Score: 1

    ATI's motherboard chipsets will be sold to OEM's under the AMD label instead of the ATI label.

    I think ATI's motherboard chipset business is what AMD really wants. This way, they don't have to wait for nVidia to come out with the latest nForce motherboard chipsets when it can be developed in-house using ATI's own technology.

    However, expect the ATI name to remain for graphics card chipsets due to the name recognition factor of the ATI brand.

    1. Re:Here's one obvious thing: by NXprime · · Score: 1

      Would you pay $5.4 billion for a southbridge? AMD already has the northbridge slapped on thier CPU.

      Oh and Intel won't renew ATI's chipset license. So no chipsets for Intel anymore. http://www.theinquirer.net/default.aspx?article=33 225

      AMD always had the capability to make thier own chipsets, they have choosen not to, that's all.

  56. Ugh by PhotoGuy · · Score: 1

    I prefer AMD over Intel. I prefer NVidia over ATI (better Linux support). This move likely means I won't be seeing any integrated nVidia on AMD boards in the near future, which truly sucks. So will Linux support for AMD boards suddenly become worse (due to ATI's stinginess on the specs), or will AMD make ATI be more liberal with its specs now? It would certainly be in its best interests to do so.

    --
    Love many, trust a few, do harm to none.
  57. Re:Intel has killed gaming...but AMD has restored by ichigo+2.0 · · Score: 3, Informative

    The problem is that Intel's integrated graphics is utterly incapable of running games, it doesn't even properly support DX9. On the other hand both Ati's and Nvidia's IGPs are DX9 compliant and lightyears faster than Intel's IGP. Thus we come to the conclusion that Intel is hurting gaming with it's crappy IGP, because you can't buy a similarly crippled IGP for AMD even if you wanted to. Luckily the next version of Intel's IGP will at least support the DX9 spec properly, as it wouldn't be able to display Vista's Aero UI otherwise..

  58. Damn? by tomstdenis · · Score: 1

    I personally think this is a stupid idea. Part of the appeal of AMD was that you didn't have just one source for your parts. I'm sure merging with ATI will piss of Nvidia [and possibly even VIA], maybe even to the point that they stop making AMD required parts. They better work hard to ensure that the customers are the ones who get to decide. Otherwise they're fundamentally no better than Intel. :-(

    Tom

    --
    Someday, I'll have a real sig.
  59. Re:See Hackers Dictionary: "Wheel of reincarnation by havoc- · · Score: 1

    This makes it sound almost like a bad thing. Why not develop some new technology seperately so it can mature, then integrate it bake into the 'big core' for better efficiency, develop a new generation seperately, etc...? Not a bad thing as I see it, but a natural cycle of technology.

  60. Oh shit, this sucks. by mobby_6kl · · Score: 1

    Incredible, one of the two major CPU companies buys one of the two major GPU companies, and slashdot is rejoiced because this will result in "better linux drivers"! Are you people nuts? The possible benefits of that are tiny compared to the havoc this deal might cause if AMD decides to play politics with its new toy.

    Imagine if Intel bought NVIDIA and decided to axe the nForce chipsets. AMD would be fucked, being stuck with crappy via (or whatever) chipsets. Of course, Intel makes its own chipsets, so it's not exactly the same, but the potential of problems is there.

    1. Re:Oh shit, this sucks. by openldev · · Score: 0

      I really don't see an Intel acquisition of nVidia in the future. As for the Linux drivers, I agree with you that it's not a _big_ deal, but to someone that wants the driver support, it can be a big deal. Although, enough comments are enough about that ... I think I've read about 500 billion comments on Linux drivers from ATI, LOL.

  61. AMD up to their eyes in debt by edxwelch · · Score: 3, Interesting

    They borrowed $2.5 billion to pay for ATI. This is top of all the other debt that they owe, they still haven't payed off the massive cost of the 2 fabs in Germany and they also own a lot of stock in Spansion which itself is heavly in debt.
    AMD has been loosing money for a lot of years (only in the last 2 years they started making profit)
    Now they have a price war with Intel and they have to compete with Conroe, so they can't even count on making any profit from the next few quarters.
    Looks like they are living on the knife edge.

    1. Re:AMD up to their eyes in debt by Anonymous Coward · · Score: 0

      Yes, it looks highly likely that AMD will return to red quarters again.

  62. Driver code not the issue by CustomDesigned · · Score: 3, Interesting
    As I understand it, the secrets GPU companies want to protect are not the driver itself, but what low level functions are provided by the card. Open sourcing the driver would expose those functions. IMHO, GPU companies should export a higher level interface from the card - say OpenGL, via firmware on the card. They could have OpenGL and DirectX ROMs for the card. Or even ship with DirectX by default, and offer an "upgrade" to OpenGL. (Not as nice for Linux folk like me, but M$ have dominant market share.) The OS driver would be a very thin layer. This would keep all their hardware secrets in the card where they belong, but allow full use under open source operating systems.

    The drawback would be a lockout for experimental 3D APIs. But it would be no worse than the binary driver situation we have now.

    1. Re:Driver code not the issue by baadger · · Score: 1

      The problem with this is you essentially move the driver's functionality into firmware, making the driver pretty pointless.

    2. Re:Driver code not the issue by dnoyeb · · Score: 1

      So they would make all of their cards more expensive and difficult to support by adding an additional layer of firmware just to support the smallest segment of their market?

    3. Re:Driver code not the issue by CustomDesigned · · Score: 1

      Exactly. The driver becomes the equivalent of the GPL compatibility layer that Nvidia ships with their binary drivers now. The binary driver moves to firmware.

    4. Re:Driver code not the issue by CustomDesigned · · Score: 1

      A valid reason why there is no motivation to do so now from a short term business perspective. But when "open source friendly" becomes valuable enough marketing wise to make them want to do it, this is just an idea for protecting hardware secrets at the same time.

    5. Re:Driver code not the issue by Panaflex · · Score: 1

      I'm all for that. What's the point of drivers anyway? IMHO, they should be simple I/O devices that know how to drive the hardware.

      --
      I said no... but I missed and it came out yes.
    6. Re:Driver code not the issue by beoba · · Score: 1

      Picture this scenario: Reformatting your computer, and not having to redownload drivers for your video card, since they're already on the card itself. All the OS needs is the 'translation layer'.

      --
      I am not a number - I am a free man!
    7. Re:Driver code not the issue by Brandybuck · · Score: 1

      Sounds good to me! I'm having problems figuring out the "problem" you refer to.

      --
      Don't blame me, I didn't vote for either of them!
    8. Re:Driver code not the issue by ChronosWS · · Score: 1

      What is the cost-benefit analysis for this? Why should they release anything or go to any extra trouble at all? An open-source driver provides essentially no benefit to them. Remember, these cards are (at release time) geared towards high-end gaming systems, not enterprise-class systems where a company might reasonably think open-sourcing the drivers could be useful (and where there is a significant market share.)

    9. Re:Driver code not the issue by Anonymous Coward · · Score: 0

      Then picture a failed firmware upgrade bricking your $700 video card. One of the nicest things about having the drivers exist outside the card is that you can increase the performance of the card by tweaking the drivers with little to no risk.

    10. Re:Driver code not the issue by powerlord · · Score: 1
      One of the nicest things about having the drivers exist outside the card is that you can increase the performance of the card by tweaking the drivers with little to no risk.


      Sounds like another reason that the GPU companies would like this approach :)
      (I'm not at all convinced they like the idea of people overclocking their chips or tweaking them instead of buying higher priced components).
      --
      This space for rent. All reasonable inquiries will be entertained at proprietors discretion.
    11. Re:Driver code not the issue by baadger · · Score: 1

      It makes it all the more difficult and risky to upgrade your hardware's interaction with software?
      Makes it more difficult for open source?
      Makes drivers less portable (to CPU architectures and platforms)?

    12. Re:Driver code not the issue by mrchaotica · · Score: 1
      IMHO, GPU companies should export a higher level interface from the card - say OpenGL, via firmware on the card.

      You know, the fear of exactly that is what drives projects like OpenGraphics and OpenBIOS. Binary blobs are bad whether they're in software or firmware.

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    13. Re:Driver code not the issue by Brandybuck · · Score: 1

      Huh? Firmware doesn't cause those problems, it makes them GO AWAY!

      --
      Don't blame me, I didn't vote for either of them!
  63. nVidia not going to cut off AMD by argent · · Score: 1

    Intel makes GPUs and nVidia still makes chipsets for Intel motherboards.

    1. Re:nVidia not going to cut off AMD by tomstdenis · · Score: 1

      What if the wording of the deal prevents AMD from licensing to nvidia for HT 3.0 or whatever in the future?

      Tom

      [Disclaimer: They don't tell me shit. So don't take this as company endorsed slashdotting...]

      --
      Someday, I'll have a real sig.
  64. i hope AMD fixes ATI's issues ;) by Intangion · · Score: 1

    ATI kinda sucks for linux support, and their drivers in general seem more unpolished and buggy

    hopefully AMD gets things moving to provide better drivers and better linux support for their supposedly superiour hardware..

    nvidia does a great job with their drivers. if intel bought them id be worried they screw it up with 'strategy' and crap, just give us the damn hardware

  65. What the...? by TouchOfRed · · Score: 0

    This could be one of the biggest mergers in computer technology, and all the comments I seem to find are OMFG ATI LINUX DRIVERS?!(not all... but you get the picture). This has nothing to do with linux. Honestly if you need the drivers THAT badly, go find a relative with a windows computer and an equally powered nvidia card and swap it. Dont buy ati products if you want to use linux. In reality, you wouldnt buy a moped to drive on the highway, so dont go bitching when it doesnt go past 80km/h on the linux highway. (Dont take this as ati products are slow, take in the "right tool for the job" lesson).

  66. ATI no longer competes directly with NVidia by Alzheimers · · Score: 3, Interesting

    Taking into account all the fanboi anguish, let me point out the very simple fact that now ATI no longer directly competes with NVidia. You could say that the competition would be between AMD and NVidia now, but that's not quite right either. The fact is that the market has become so diverse that all these companies were already competing with each other, despite partnerships and deals.

    AMD, ATI, NVidia and Intel *all* make motherboard chipsets.
    ATI, Nvidia and Intel all make video processors.
    So do SIS, S3, and VIA.

    Yet they all work (relatively) well with each other.

    This isn't about marketshare, it's about technology. ATI does something that AMD wants, so AMD is acquiring the company for the tech. The market won't feel a thing, I promise you. Competition will continue, just like it did when Micron acquired Rendition (wipes a tear for his Verite v2200) and when NVidia bought out 3dfx (wipes another for his Banshee).

    Since everyone's got their prognosticator's caps on today, I'm going to come out and say that, within 5 years, we'll be seeing GPU processors integrated into the motherboard, accessable to both ATI and NVidia (and Matrox, and S3, and ...) The power and bandwidth demands for next gen GPUs are becoming more than expansion boards can handle. Instruction sets are becoming extremely CPU-like. Since the whole universe seems to be moving into Multi-processor designs anyway, perhaps we'll even see some kind of GPU-MMX style expansion of the x86 instruction set (call it v86 for now).

    I think we're seeing a move back to specialization. We've already got separate Audio chips, separate networking chips, even chips to handle I/O for RAID and such. With the new market for Physics co-processors, I'm sure we'll only see more for tasks such as AI, and when the next big UI design is unleashed (either some kind of brain-reading technology, or a true 3d input system -- the WII is just the tip of the iceberg!) another co-processor will be made to handle that. With AMD's focus on integrating external processors with technologies such as HyperTransport, undoubtedly they'll be able to compete for a long time.

    And the best part is, we get to choose from strong market competitors. As long as there is innovation, we win.

    1. Re:ATI no longer competes directly with NVidia by Anonymous Coward · · Score: 0

      Everyone is commenting about how this is just to acquire technology. AMD has stated that they're developing HTX slots to turn almost any high bandwidth chip into a coprocessor with a direct link to all other processors. You can't get much higher bandwidth with a need for low latency for cpu instructions than graphics. Add to that the current work on GPGPUs and I wouldn't be surprised if we saw ATI on Hypertransport bus in the future.

    2. Re:ATI no longer competes directly with NVidia by GnuPooh · · Score: 1

      I think your information is dated. AFAIK,

      AMD doesn't make chipsets currently.

      S3 market existance is questionable. NewEgg doesn't sell anything and I can't find anything at PriceWatch. Froggle had a few places, but I think this is just old stock.

      SIS lists all their video chips as discontinued.

      VIA's graphics are S3.

      So it really is fair to say that the only external graphics chip makers are ATI and NVidia and Intel does chipset graphics. Don't make it seem like there's more competition that there really is currently.

    3. Re:ATI no longer competes directly with NVidia by usrusr · · Score: 1

      > I'm going to come out and say that, within 5 years, we'll be seeing GPU processors
      > integrated into the motherboard, accessable to both ATI and NVidia (and Matrox, and
      > S3, and ...) The power and bandwidth demands for next gen GPUs are becoming more
      > than expansion boards can handle.

      I really did not see this coming until after the merger, but now i realized that the maths is quite simple:

      Back in the days of 3dfx (and, of course, earlier) the speed from cpu to video DAC was a bottleneck that had to be overcome by painting to framebuffer with an accompanying RAMDAC and trying to generate as much of the framebuffer content updates by some external chip (the GPU).

      In the days of digital displays we don't even need a video DAC anymore and at the same our CPUs are starting to come with an external bandwidth that could easily carry multiple DVI links (a quick check at wikipedia rates hypertransport at roughly 20 GByte/s while dual link DVI is rated less than 1 GByte/s). To make matters even more clear, the current ways of externally generating the content of the framebuffer often demand more bandwidth for communicating between CPU/RAM and GPU (AGP 8x is listed as 2 GByte/s, 16 times as much as dual link DVI) than for moving the result down the DVI 60 times per second.

      Conclusion: the bottleneck has moved, once it made sense to move as much of the image creation as close to the RAMDAC as possible, today everything looks like moving the image creation as close to CPU/RAM as possible is the thing to do. As it happens, in the AMD world CPU and RAM are effektively at the same place.

      Today we have very fast RAM at the GPU end of things, but moving that closer to the CPU, acting as another caching stage for both general purpose and the moved graphics units could only make things faster.

      Add to that the fact that todays GPU pipelines have little in them that is strictly graphic specific and that external number crunching chips (that seem to have a lot in common with those still-GPUs) seem to be taking over both the entertainment PC and the high performance clusters and you get a glimpse of an x86 future that looks as surprising and strange as it looks promising.

      --
      [i have an opinion and i am not afraid to use it]
  67. Support for linux to improve? by Anonymous Coward · · Score: 0

    Will AMD's oft advertisied, and frequently proved, support for linux, drive the new ATI to support linux in the same timely manner as 'doze?

  68. Graphics in software by gr8_phk · · Score: 2, Informative
    Processors are getting fast enough to do rendering in software again. GPUs are trying to become general purpose CPUs. People will soon have 2 cores as standard and 4 or more are on the way. What are people supposed to do with all those CPU cores? Replace the GPU with them of course. Why would a CPU maker need to buy a GPU maker? Not sure, but perhaps just to gain the graphics expertise to write the software, and possibly to make some suggestions for the instruction set and hardware. I certainly hope they don't just integrate the GPU onto the same die - that's so unnecessary.

    IMHO a much better - and cheaper - buy would have been that company (don't remember name) that has a really fast DirectX implementation in software. You don't even need a frame buffer any more, just put some circuitry in the north bridge that pulls data from main memory and spits it out over DVI-D. Let software on multiple cores take care of rendering.

    For reference, I think we're still on track for software realtime raytracing by 2012. If we can do that, certainly software can rasterize polygons fast enough before then. The GPU as we know it is dead - good for ATI finding an exit strategy, bad for AMD spending so much money. What's nVidia's exit strategy? Intel already knows a lot about graphics.

    1. Re:Graphics in software by p0tat03 · · Score: 1

      Getting rid of GPUs and getting CPUs to replace them is a terrible idea. CPUs are designed to general-purpose computing, while GPUs are able to exploit the embarrassingly-parallel nature of graphics rendering. That 650MHz GeForce 7900GTX gives you a lot more oomph than a 3GHz P4 ever will. I don't mind integrating GPUs onto CPUs, but you can't just REPLACE them wholesale with general-purpose CPUs.

    2. Re:Graphics in software by gr8_phk · · Score: 1

      You miss the point. In 6 years we will be doing ray tracing on the CPU(s) in real time. That's better graphics than GPUs can do. At that point, the GPU is truely redundant. When is the last time anyone tried to do rendering on the CPU? Quake 1? It's come a long way since then, and now we've got extra cores that no one knows what to do with. Besides, for the majority of business computers a GPU is already unnecessary - windows can be painted in software, even with fancy compositing.

    3. Re:Graphics in software by drinkypoo · · Score: 1

      Unreal and Unreal Tournament (same engine, more or less) have/had a software renderer.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    4. Re:Graphics in software by smallfries · · Score: 3, Interesting

      Ok, time to blow off the moderations that I've made so far as you're missing something fairly obvious. Take a look at the processor on the 7800GTX, with 300 million transistors it is currently the most complex chip being shipped (I don't how big Cell is). In exchange the peak processing power is 320Gflop/s (40 in the vertex processors, 280 in the fragment shaders). For comparison the floating point performance in a CPU is ~ 8Gflop/s. That's a whopping 40 cores to break even on graphics processing.

      Once you can fab a processor large enough to contain 40 functional cores - how big a GPU do you think you could fab on the same process? The simple fact is that a GPU is completely crippled compared to a CPU. There are huge tradeoffs in the design to get that kind of performance. Stream processing is very limited compared to a von Neumann architecture if you care about latency in the slightest. But for graphics - it's perfect. Throwing completely independent parallel chunks of data through an array of vector processors is a much simpler challenge than attempting to extract parallelism from sequential code. The sequential code has pesky things like control-flow that is missing in the gfx shaders, and I don't mean the rubbish that ATi/Nvidia are selling as control-flow in their current designs. That is sheer marketing given the size of the shader batches and the depth of the pipelines.

      So I don't think the big 'ol wheel of reincarnation is going to move rendering back into software anytime soon. But what people forget is that AMD is not really a processor company. They are a fab company that just happens to design some kick-ass processors. Their main business is silicon, and buying ATi is the biggest chunk of vertical integration you can imagine...

      --
      Slashdot: where don knuth is an idiot because he cant grasp the awesome power of php
    5. Re:Graphics in software by John+Courtland · · Score: 1

      Perhaps the idea is to find a way to architect multiple GPU-like cores into looking like a current single core. AMD did state that they were looking at making multi-cored chips appear to the host as 1 CPU. Perhaps this is part of that idea?

      --
      Slashdot is proof that Sturgeon's Law applies to mankind.
    6. Re:Graphics in software by usrusr · · Score: 1

      add some of that ultra-fast memory from the GPU to the CPU as some kind of level-whatever cache, and you get closer.

      then, realizing that graphics work can be parallelized like flies you then replace half of your complicated out-of-order-execution general purpose cores with a number of smaller cores that each get half the bang out of 1/10 the transistors. those would still be general purpose, but there is so little left in today's GPUs that is really specific to a graphics pipeline, that it would not make much of a difference. still, those cores (or mmx style special units) would be much easier be brought to use for purposes other than gaming than today's GPUs.

      we agree that the basic nature of graphics processing lies in the parallelism. why limit your parallelism-oriented resources to graphics processing?

      --
      [i have an opinion and i am not afraid to use it]
    7. Re:Graphics in software by arose · · Score: 1

      What about additional instructions for graphics rendering?

      --
      Analogies don't equal equalities, they are merely somewhat analogous.
    8. Re:Graphics in software by smallfries · · Score: 1

      The big distinction is how much parallelism exists in the workloads. The classic quote in the literature for graphics is that 'It is an embarassingly parallel task'. Most general workloads are not. I think the main dead with the multi-core -> single-core work was that it was targetted at multi-threaded code. You can't really fake a single core that well for a sequential application because of the latency, but you can hide latency really well across multi-threaded code. The problem with building a single sequential core out of cheap vector parts is the same.

      I think someone further down the page has the right idea - imagine you are building a 4-core processor (like AMD). Throw one of the cores away. Use those transistors to build a mini-vector processor. It's not that graphics specific. You get more raw Gflop/s for the 'right' applications from that one core, and a lot more processing for the 'wrong' out of the other three than out of any hybrid architecture. Suddenly its a new tradeoff - how much vector performance, how much scalar?

      --
      Slashdot: where don knuth is an idiot because he cant grasp the awesome power of php
  69. Hmmmm, Consoles by MrCopilot · · Score: 3, Insightful
    Q.) How many next-gen consoles have AMD in them now?

    A.) Xbox, Nintendo

    Analysis.....Good move.

    --
    OSGGFG - Open Source Gamers Guide to Free Games
    1. Re:Hmmmm, Consoles by dreamchaser · · Score: 1

      It has zero to do with consoles. Consoles do not make any significant money for chip makers. Consoles make money for game publishers.

      It has *everything* to do with wanted ATI's technology and intellectual property, for reasons already covered elsewhere in other people's comments.

    2. Re:Hmmmm, Consoles by Ichigo+Kurosaki · · Score: 1

      Accually those are IBM cpus...
      and sony has a IBM chip as well.
      Cheers

    3. Re:Hmmmm, Consoles by ScaryFroMan · · Score: 1

      No, he's referring to the ATI GPUs, not the CPUs, which are IBM.

      --
      In Soviet Russia, backwards is everything.
    4. Re:Hmmmm, Consoles by MrCopilot · · Score: 1
      It has zero to do with consoles. Consoles do not make any significant money for chip makers. Consoles make money for game publishers.

      Traditionally Consoles make no money for Console Makers. (Nintendo Excluded)

      However the chips inside cost money, Ati was happy to have a GPU in the GameCube, and They (AMD & Ati) are happy to be in the Wii and the 360. I have no numbers but I'd bet my next console (Wii) they didn't sell below cost.

      --
      OSGGFG - Open Source Gamers Guide to Free Games
  70. Nvidia by larryau · · Score: 1, Funny

    They should have bought Nvidia instead.
    Nvidia Fan :)

  71. Re:See Hackers Dictionary: "Wheel of reincarnation by ivan256 · · Score: 1

    Nah, it's not a bad thing. It's just fairly reliable and easy to base predictions off of. Personally, I like to bring it up anytime somebody is insisting that the course of technology for a particular market is *perfect* and should continue marching forward unchanged.

  72. Pfficial? by skyh0rse · · Score: 1

    It's Pfficial: AMD Buys ATI Pfficial?

    1. Re:Pfficial? by Anonymous Coward · · Score: 0

      Pwned!

  73. NVIDIA can't open up drivers by Anonymous Coward · · Score: 2, Insightful

    NVIDIA has stated many times that even if they wanted to open up documentation to their cards, they can't.
    There are cross-licensing issues that prophibits them from releasing the specs.
    So they release closed binary drivers for linux instead.

    Stop whining people, they are doing the best they can.

    (P.S I can imagine ATIs situation is similar)

    1. Re:NVIDIA can't open up drivers by makomk · · Score: 1

      Stop whining people, they are doing the best they can.

      Not really. The Linux driver I'm using for my NForce network card? It's reverse-engineered. They were more than willing to release a binary blob for Linux, but didn't want to release any specs. (Oddly, once the open-source driver was finished and popular, NVidia were more than willing to contribute to it, though I don't think they ever released any specs at any point...)

  74. AMD going into embedded devices by chris_7d0h · · Score: 2, Interesting

    According to the merger telco. the only substantial argument for the merger from the company's side is that they want to get into the embedded device business. They hope to provide a platform for media processing on cell phones, TVs and the like.

    The Q&A session is apparently already up at The Pirate Bay (though I didn't manage to download it yet): http://thepiratebay.org/details.php?id=3506714

    Interesting that they think they'll be able to continue having a good relationship with nVidia. I'd guess it's just PR speak though for "as soon as the merger is complete, you're unimportant to us".

    The CEO Hector Ruiz went on and on like a drone, repeating the same fluff over again (like background noise) and it wasn't until those few moments where his minions were allowed to speak something intelligible was said.

    --
    In a society that believes in nothing, fear becomes the only agenda ~ Bill Durodié
  75. Better ATI drivers... by HavokDevNull · · Score: 2, Interesting

    Wonder if AMD will force ATI's hand to Finlay release a decent ATI video driver for Linux now?

    --
    Sig
    1. Re:Better ATI drivers... by brain1 · · Score: 1

      ... we can only hope. Maybe then I might consider ATI hardware. But not until I see an actual, working Linux driver from them.

    2. Re:Better ATI drivers... by Elf-friend · · Score: 1

      One can hope. AMD should certainly be awrae of the strong support they have amongst technically astute users, espicially the *nix crowd (BSD as well as Linux). If ATI released proper *nix drivers, I'd drop nVIDIA in a heartbeat.

  76. Re:Intel has killed gaming...but AMD has restored by guruevi · · Score: 1

    The problem is that Intel does support DirectX8 and 9 (depending on the chip). The issue with it is, it doesn't accelerate, it let's the processor take over it's work and it doesn't support OpenGL.

    --
    Custom electronics and digital signage for your business: www.evcircuits.com
  77. I wonder by OfNoAccount · · Score: 1

    Would it be possible for ATI/AMD to produce GPUs for the latest CPU sockets (AM2/SoF)? That way you could just configure a system based on the performance you need - if you need more GPU power just load the sockets up with GPUs instead of CPUs. I know one company already makes AMD socket co-processors, any reason this couldn't be done with GPUs too?

    Obviously you'll need somewhere to actually dump the video signal, but there's no reason that I can see for not providing a DVI socket on the mobo?

    You'll also need some nice fast memory, but if you have a fair amount of on-die cache you could mitigate problems I suspect. HT would be better than PCIe/Crossfire/SLI from a bandwidth perspective wouldn't it?

  78. The circle is now complete by dino213b · · Score: 1

    Some of the reader comments are starting to make sense. We are headed for a full circle. For example, we've had this happen:
    CPU -> separate FPU -> CPU w/FPU evolution

    Now it seems like we have a:
    CPU -> separate GPU -> CPU w/GPU evolution

    What's next? Probably a:
    CPU -> physics processing unit (PPU) -> CPU w/PPU evolution

    So following this design trend, are we to assume that every design will come back to a mutated square one? It seems like we should have done this from the start, skipping the middle steps. Of course, in the real world, things work differently.

  79. hopefully better support by sabernet · · Score: 1

    I have an ATI card and their remote wonder II. Their hardware is awesome. However, their software just plain blows. Their website is horrid and their support is disgusting. I got my card with a currupted driver disc. As there are things in there you couldn't download from the ATI site, I tried to get ATI to send me one. After being hung up on 3 times, calling an extra 4 times(note, this stuff ain't toll free), getting automated email responses etc... I finally got the damn CD.

    On top of that, the RWII was a mail in rebate that came with the AIW card. Four months later, another nightmare because they had misplaced my application and receipt even though I tracked the letter straight to their offices.

    Their products manuals blow, their site is uninformative, they have no linux drivers, they have no stereo drivers....yaaaaargh. I really hope this buyout causes some change in that regard.

  80. Re:See Hackers Dictionary: "Wheel of reincarnation by Anonymous Coward · · Score: 0

    Hello specialized Physics processors:

    see: http://www.ageia.com/

  81. Think about what you just said by default+luser · · Score: 5, Insightful

    Now tell me: why are the likes of NVidia and ATI keeping their products undocumented and their drivers closed?

    Because, if they DO PROTECT THEIR IP, The OTHER GUY has to waste TONS OF MONEY on reverse-engineering teams and highly-qualified people to reverse-engineer the processor via electron microscopes.

    It's not the EQUIPMENT that is expensive, it is the PEOPLE. And, as you Linux zealots know FULL WELL, reverse-engineering is EXPENSIVE in terms of PEOPLE and TIME.

    If you publish the specifications of your latest graphics chip for all to see, suddenly your competitors don't have to divert staff from working on next-generation architectures just to reverse-engineer your system. Instead, they can analyze your documentation in a fraction of the time.

    It's a two-way street, so stop deluding yourself that there's only one side to the story. Publishing full specs for your graphics chips is like writing your competition a blank check. Intel is the only one who doesn't have an issues doing this because their graphics technology is always following.

    And to counter your argument: what happens in two years when ATI and NVidia decide your card is too old to support, and yet it still performs very well but you NEED the features in the latest kernel and latest x.org? Go ahead, buy a new video card.

    Yes. There are still many well-supported video cards sold in AGP. In fact, you can still get well-supported video cards in PCI, a fifteen-year-old technology. They're not top-performers, but beggars can't be choosers.

    The video card market is transitioning to PCIe with surprising speed precisely because they do not want another VLB fiasco. The PCI -> AGP transition was slow because PCI still had a future for other types of cards, but the AGP -> PCIe transition was rushed to avoid market confusion. You can still buy plenty of AGP cards, but the big players have made it clear: there won't be any more improvements for AGP.

    --

    Man is the animal that laughs.
    And occasionally whores for Karma.

    1. Re:Think about what you just said by Anonymous Coward · · Score: 0

      You need an electron microscopes to reverse engineer a CPU, not a graphics card. They're made up of three parts: RAM, a general purpose FPU, and FPGA's for the vector ops. There might be some interesting designs with respect to bus interconnects, and that sort of thing will show up under a plain old magnifying glass. How the architecture is used will be revealed in the drivers, but the drivers also contain the really proprietary stuff, namely the configuration information for the FPGA's.

      So no, it's not true that the drivers "just talk to the hardware" and that all the really proprietary stuff is in the hardware. These days, the driver is the hardware, or at least the firmware microcode at any rate.

      Still, it'd be nice if they could at least factor out a little more open source from the drivers than the monolithic blobs they are now. But that's not how they design their software.

    2. Re:Think about what you just said by Lisandro · · Score: 4, Insightful

      If you publish the specifications of your latest graphics chip for all to see, suddenly your competitors don't have to divert staff from working on next-generation architectures just to reverse-engineer your system. Instead, they can analyze your documentation in a fraction of the time.

          Bullshit, sorry. We don't want their beloved silicon blueprints of their latest GPUs, just information on how to make them work. Want to draw a polygon? Send this command to the card. Do a hardware T&L? This other one. You can learn only so much from a driver sourcecode or techincal specifications on how to program a GPU. Don't beleive me? Check the information released by both nVidia and ATI for their older GPUs, and see how much you can infer from them.

    3. Re:Think about what you just said by Fulcrum+of+Evil · · Score: 1

      It's not the EQUIPMENT that is expensive, it is the PEOPLE. And, as you Linux zealots know FULL WELL, reverse-engineering is EXPENSIVE in terms of PEOPLE and TIME.

      I like TO randomly CAPITALIZE words. I think it MAKES my POINT sound VALID.

      --
      "We returned the General to El Salvador, or maybe Guatemala, it's difficult to tell from 10,000 feet"
    4. Re:Think about what you just said by dvice_null · · Score: 1

      > If you publish the specifications of your latest graphics chip for all to see Ok, but would it be possible to publish specifications for older chips?

    5. Re:Think about what you just said by Cocoshimmy · · Score: 1

      Well Said!!

      I don't think a company should sacrifice their IP just to appease a small, but vocal, linux user base. Especially seeing as most gamers use windows anyways and it would give the competition that extra advantage. People just don't understand how competitive the graphics card industry is. It's probably the most competitive semiconductor industry out there, more so than the CPU business.

    6. Re:Think about what you just said by vertinox · · Score: 1

      Because, if they DO PROTECT THEIR IP, The OTHER GUY has to waste TONS OF MONEY on reverse-engineering teams and highly-qualified people to reverse-engineer the processor via electron microscopes.

      Hrm... That sounds like wearing a bullet proof vest to force the other guys to drop an expensive missile on your head to kill you.

      The only problem I see with this is that you are still dead, but if the fact that just dropped an object that costs $25 million dollars on you is any consolation... So be it.

      --
      "I am the king of the Romans, and am superior to rules of grammar!"
      -Sigismund, Holy Roman Emperor (1368-1437)
    7. Re:Think about what you just said by Anonymous Coward · · Score: 0

      The video card market is transitioning to PCIe with surprising speed precisely because they do not want another VLB fiasco. The PCI -> AGP transition was slow because PCI still had a future for other types of cards, but the AGP -> PCIe transition was rushed to avoid market confusion. You can still buy plenty of AGP cards, but the big players have made it clear: there won't be any more improvements for AGP.

      BULLSHIT. You're naive if you think they're cutting off AGP for our own good, they're doing it to sell more motherboards, to get people to make big upgrades instead of small ones, to encourage SLI and to be able to gouge on AGP cards.

    8. Re:Think about what you just said by default+luser · · Score: 1

      Exactly. It's like MAD, except there's no threat of personal injury to the people involved. Each side is worried that if they give so much as an inch, the other side will take two more.

      Sure, it doesn't make sense, but it doesn't have to; that's just the way escalalation works out.

      --

      Man is the animal that laughs.
      And occasionally whores for Karma.

    9. Re:Think about what you just said by goaty_the_flying_sho · · Score: 1

      What about publishing the specs for your new ethernet controller? This isn't the 1980s, why did nVidia force two people to cleanroom reverse engineer the forcedeth driver?

    10. Re:Think about what you just said by Anonymous Coward · · Score: 0
      Because, if they DO PROTECT THEIR IP, The OTHER GUY has to waste TONS OF MONEY on reverse-engineering teams and highly-qualified people to reverse-engineer the processor via electron microscopes.

      It's not the EQUIPMENT that is expensive, it is the PEOPLE. And, as you Linux zealots know FULL WELL, reverse-engineering is EXPENSIVE in terms of PEOPLE and TIME.

      PLEASE stop capitalizing RANDOM words IN YOUR sentences. You AREN'T adding EMPHASIS, you're JUST making it HARD TO read.
  82. One correction. by LWATCDR · · Score: 3, Informative

    "AMD doesn't like closed technology like Intel does. So it'll be an open platform still which is a 'good thing' (tm). "
    Actually Intel has been a big supporter of OSS. They helped port Linux the Itanium and have provided all the documentation to their video chips.
    I think you are confusing Intel with Microsoft. Intel has been one of the most open hardware companies.
    AMD has also been very good. ATI like nVidia.... Well let's say not so good.
    I really don't get this.
    AMD could use some good chip-sets but they have made their own for the Opteron so I don't see what they gain from ATI.
    AMD could use a good low end integrated video solution for low end desktops and servers. Yes it is true but servers almost never use nVidia or ATI graphics cards. When I set up a server I only plug the monitor in when I do the install and if something really bad happens.
    I have to think this comes down to laptops. AMD has not done well in that market and a one stop shop for a laptop solution like Intel offers might be a good solution.
    I wouldn't hold my breath on the good open source ATI drivers for Linux. Of course if it happens I might dump my nVidia based motherboard and Video card. I have been buying nVidia just because of their better Linux support for years.

    --
    See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    1. Re:One correction. by Anonymous Coward · · Score: 0

      Um, you should really look at 1U, 2U, 4U, & 6U servers from 1st tier manufacturers. Servers from the big boys almost always uses an integrated ATI chipset on the motherboard.

      Thing is it's invariably an ATI RageXL or equivalent - something that would be laughed at in the consumer or even business market. By way of comparison a RageXL is less powerful, from a 3D accelerator standpoint, than a 3Dfx Voodoo3. For 2D, assuming you have extensive driver support, ATI is faster than a V3 - on top of which ATI has their ever-present treasure trove of "multimedia" acceleration (offload MPEG from the CPU & similar whizbang cool stuff).

  83. Re:Intel has killed gaming...but AMD has restored by MrSquirrel · · Score: 1

    If gamers want to game, they can splurge and buy a graphics card (a decent card is just as expensive as 2 of the latest games ~$120). I don't disagree that Intel integrated graphics are utter garbage, but it's not hard to circumvent poor onboard graphics with an AGP/PCI-e card. Even if there was a motherboard with amazing integrated graphics, I would pass it up because in another 2 years I would have to buy a whole new motherboard (with probably a different CPU socket, so a new CPU would be needed too) with the then-latest in gfx chips in order to play the latest games.

    --
    A computer once beat me at chess, but it was no match for me at kick boxing.
  84. ATI are a complete joke ! by Anonymous Coward · · Score: 0

    See title, what more needs to be said?

  85. And laptops. by LWATCDR · · Score: 1

    I agree about the low cost integrated video chips but I have to add one thing.
    Laptops.
    AMD really lags behind Intel in the laptop market. With ATI on board they could offer a complete AMD laptop solution like Intel does.
    You also left out the other place that low cost integrated video chips live, servers.
    AMD does care about that market.
    Now I have to wonder if we will see an integrated video solution from AMD that uses Hyper-transport to talk to the onboard video?
    That could give AMD a leg up in the HD video market.
    Lots of interesting things could happen.

    --
    See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
  86. The simple truth by ravyne · · Score: 1

    The truth is that AMD needed to bring on a solid chipset and integrated GPU in-house. Intel has been moving in the direction of providing the whole package, most importantly CPU+Chipset+GPU that form the backbone of the system. By purchasing ATI, AMD has brought Chipsets and GPUs into their fold and can compete directly against intel's platform strategy.

    ATI's Integrated GPUs are light years ahead of Intel's crap, Even though the specs look better on paper for intel's part, real world performance is another matter entirely. Game developers simply don't care whether or not their game even runs on intel's GPU, let alone at an acceptable rate. ATI, because their integrated chips are derived from their discreet parts, does not have this problem. ATI's current integrated chip is simply a two-pipelined version of the same architecture as its X800 GPU with their hyper-memory technology and, optionally, a 64bit bus to add dedicated graphics ram. It may not provide the best gaming experience, but it will run any modern title you throw at it.

    I don't expect to see AMD integrating a GPU into their CPUs any time soon. It would add too much to the transistor count and the turnover in GPUs is too quick to be all that usefull. However, AMD has been pushing their multi-socket configuration as a strategy for the future and have opened up the hyper-transport spec to outside firms so that they can develop ASICs that plug directly into an AMD-based system. Its entirely possible that we may see a GPU in this form.

    1. Re:The simple truth by tomstdenis · · Score: 1

      Psst, AMD has their own chipsets for AMD processors... what do you think others use as reference? What this brings to the table is a GPU that AMD can pwn.

      Tom

      --
      Someday, I'll have a real sig.
  87. Another good Canadian company bites the dust... by Kenshin · · Score: 1

    Sooner or later americans are gonna own EVERYTHING up here.

    Not only that, they kill everything they touch. (Just ask everyone I know whose companies were bought by americans.)

    --

    Does it make you happy you're so strange?

  88. Don't mess with my video cards !!! by Phantom_24 · · Score: 1

    If these morons mess with my ATI card's and I can't use them on my Intel boxes, I'm jumping to nVidia !!

    1. Re:Don't mess with my video cards !!! by Elf-friend · · Score: 1

      If that happens, it will be because of Intel, not AMD/ATI.

  89. SGI? by Inoshiro · · Score: 1

    They're in the process of finishing their death throes. Will this still be an issue when they're gone in 18 months? I hope not.

    --
    --
    Internet Explorer (n): Another bug -- that is, a feature that can't be turned off -- in Windows.
  90. AMD + ATI = Intel + nVidia by Agon+Oruci · · Score: 1

    Knew this was going to happen. Now what happens to the AMD/nVidia relationship. It obviously gets broken because now AMD has ATI to make their motherboards and chipsets. Now nVidia will probably merge with Intel. If they don't nVidia beautiful drivers will die because ATI is going to have the edge in grafix cards. They will get new tech from AMD and improve their drivers like crazy. Also ATI will have super advanced technology that lets data from CPU to GPU travel 10x faster than now. That's why AMD + ATI are merging. Intel will merge with nVidia, because nVidia will not be selling sh*t and then they will be so happy to even get a deal from intel. nVidia will also be more than happy working with the Core 2 Duo. nVidia is also going to be making so much money since their grafix cards thar are in the PS3. Which will sell like crazy. So it's really going to end up like this. AMD/ATI vs Intel/nVidia = Intel/nVidia because you will have the core 2 Duo(best processor) and kik-but nVidia grafix, and nVidia will inherit the same technology that ATI is going to use to have fast travel speeds between CPU to GPU.

  91. Where are you getting the $125m? by Spaceman40 · · Score: 1

    I didn't catch that in [my translation of] the press release - just the $75m.

    --
    I [may] disapprove of what you say, but I will defend to the death your right to say it.
    1. Re:Where are you getting the $125m? by rfunches · · Score: 1

      It came off DJ Newswires:

      Rivet acknowledged AMD will lose revenue ATI gets from customer Intel Corp. (INTC), but said the combined company is confident it will make up those lost sales. AMD expects savings of $75 million by the end of 2007 and $125 million in savings in 2008, Rivet added. CEO Ruiz noted that customers were asking AMD to do this deal.
  92. maybe not niche then, but how about now? by slew · · Score: 1

    Okay, here's what I know about the current AMD...

    CPUs (okay they're still doing this one)
    FLASH (spun out that business to Spansion w/Fujitsu, did IPO a couple years ago)
    SRAM (they end-of-lifed all their discrete SRAM parts in 1999)
    PLDs (spun out that business to Vantis, sold to Lattice in 1999)
    Embededded Processors (sold Alchemy to Raza this year, Geode up for sale now)
    Microcontrollers (I think AM186 devices are "not recommended" for new designs)
    Ethernet Controllers and PHYs (although they still do sell these, mostly in niche markets)

    I think you are remembering the "old" AMD, not the "new" niche-AMD we know now...

  93. And that's the reason... by Spaceman40 · · Score: 1

    ...that being open source-friendly is so important: The geek cred.

    No, seriously. Tell your friends :)

    --
    I [may] disapprove of what you say, but I will defend to the death your right to say it.
    1. Re:And that's the reason... by dfjghsk · · Score: 1

      the 'geek cred' is very important.. this is why Dell was never able to sell its computers to anyone.

      --
      Help me take back Slashdot. When did 'News for Nerds' become 'FUD and Conspiracy Theories for Extremist Nutjobs'?
  94. ATI by Agon+Oruci · · Score: 1

    Also ATI's hardware is actually better than nVidias its just they have no drivers. They will work on that now for sure.

  95. GPU + CPU = ? by Anonymous Coward · · Score: 0

    i work at ATI. in the discrete ASIC front-end division (i.e. GPU)
    you guys obviously have no idea what you are talking about when you speak of GPU + CPU
    did you even think about the power dissipation of such architecture?
    we are already proving the limit of power dissipated per die area
    putting two chips together is not even remotely feasible at this point

    1. Re:GPU + CPU = ? by drinkypoo · · Score: 1
      did you even think about the power dissipation of such architecture? we are already proving the limit of power dissipated per die area putting two chips together is not even remotely feasible at this point

      So in a world where we have CPUs with up to eight cores (from Sun) and regularly have processors with two (intel, IBM, and everyone else these days) and where the PC is expected to go to quad-core processors next year, you're saying we can't put one CPU core and one [potentially somewhat limited] GPU core in a single package, for budget systems?

      I think you're full of shit.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    2. Re:GPU + CPU = ? by Slaimus · · Score: 1

      The GPU will also need to run much slower than the CPU like the memory controller. The potential for hotspots will likely make this unfeasible.

    3. Re:GPU + CPU = ? by The+Warlock · · Score: 1

      It's not a matter of multiple cores, it's a matter of a GPU core needs much quicker access to its RAM than a CPU core does. Of course, that problem is (relatively) easily surmountable, we just need a memory bus faster than HyperTransport. Nothing we'll be getting within the next couple of years, but five or ten years from now, everything might be on one chip.

      --
      I've upped my standards, so up yours.
    4. Re:GPU + CPU = ? by drinkypoo · · Score: 1
      It's not a matter of multiple cores, it's a matter of a GPU core needs much quicker access to its RAM than a CPU core does. Of course, that problem is (relatively) easily surmountable, we just need a memory bus faster than HyperTransport.

      What does HT have to do anything? I specified that it would need its own memory bank. However, the new hypertransport that's coming is probably going to be enough for lower-end integrated GPUs, somewhere between the intel chips and the low-end chips out from reputable manufacturers today. I do realize that the pin count of something like this would be pretty intense, but postulating a chip with a single-channel controller for the CPU side, there should definitely be enough pins to do the same on the GPU side.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  96. x86-64 is not part of the IP sharing by default+luser · · Score: 4, Informative

    This is because x86-64 is an open standard. AMD released it as open when they announced it, because it was the only way to gain industry acceptance.

    Once AMD got Microsoft's cooperation building support for x86-64 into Windows, they hardped on about the open standard. This protected AMD from Intel, who were already secretly working on their own implementation of x86-64. Normally, once Intel realized how potentially powerful x86-64 was, they were sure to create their own incompatible version (ala SSE and 3DNOW!) to try and derail AMD.

    But the open standard stopped Intel from doing this. Microsoft pointed to the open standard, and told Intel flat-out that they were not going to support two versions of 64-bit x86.

    x86-64 is an open standard. AMD's copyrighted implementation of x86-64 is called AMD64. Inte;'s copyrighted implementation of x86-64 is EMT64.

    --

    Man is the animal that laughs.
    And occasionally whores for Karma.

    1. Re:x86-64 is not part of the IP sharing by Khyber · · Score: 1

      You seem knowledgable, so perhaps you can answer this. x86-64 is pretty much a 32-bit processor (all the old x86 instructions) with 64-bit extensions, right? After all, the same 32-bit instruction set is there on the chip, and so's MMX/SSE/etc. Why is it that we can still run 32 bit programs in WinXP-x64 but we can't use 32-bit drivers, when those drivers would still work with the same damned registers that reside in the processor, and presumably the OS since it's just the same hashed 32-bit XP with the added capability to do things in 64-bit space? Where did Microsoft go so absolutely wrong that I can run 32-bit programs, but not use 32-bit drivers? What's so hard about it, if the same instruction set is there, and the OS hasn't changed all that much at it's basic levels?

      I only ask because I'm sick and tired of Windows not being able to use 32-bit drivers, but I can tell Linux with a nice --force-architecture to install and use the damned driver anyways since it's compatible with the platform it's running on.

      --
      Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
    2. Re:x86-64 is not part of the IP sharing by Anonymous Coward · · Score: 0

      There's some translation work happening in ring 3 (the application side) when you run Win32 apps on a Win64 box. You wouldn't want to impose those costs on the ring-0 drivers.

    3. Re:x86-64 is not part of the IP sharing by default+luser · · Score: 1

      I'm not sure how Linux is doing it, but Windows x86-64 runs 32-bit applications in WOW (Windows on Windows). I'm guessing Microsoft's goal was a clean 64-bit implementation with a 32-bit legacy mode for applications only, since this would by-far be the least tedious to design and support.

      --

      Man is the animal that laughs.
      And occasionally whores for Karma.

    4. Re:x86-64 is not part of the IP sharing by nuzak · · Score: 2, Informative

      You can't mix 32-bit and 64-bit modes within a single context, and drivers run in-process (with the kernel). For the same reason, you can't link 32-bit libraries in 64-bit programs, thus the reason you have no Flash plugin on a 64-bit Firefox.

      Windows used to use some really moby hacks with thunks to get 16-bit libraries working with 32-bit code, but they don't use it for NT, and opted for virtualization (WOW/NTVDM) instead. It's not perfect virtualization, but it's enough to count. Presumably they do the same thing for 32 bit code in Win64.

      It's another great argument for userspace drivers, since they could be as 32-bit as they wanted to be.

      --
      Done with slashdot, done with nerds, getting a life.
  97. You may jest... by Phil+John · · Score: 1

    You may jest, but it was probably a viable option at one point. NVIDIA already produce their own chipsets, plus they've recently purchased ULI who also make AMD chipsets.

    However, maybe AMD decided ATI could make better business orientated chipsets (NVIDIA is more "enthusiast" based and ULI have cornered the "cheap and cheerful" end of the market) and that was the reason for it. The long and the short of it is that AMD needs to produce their own chipsets for their server based parts, a-la Intel. Buyers for big business need to know everything is going to work well together.

    --
    I am NaN
  98. This SUCKS by Anonymous Coward · · Score: 0

    I want AMD + Nvidia, especially on my laptop. I don't like Intel and don't like ATI. So, soon I won't be able to buy a laptop I don't hate. Bastards.

    1. Re:This SUCKS by drinkypoo · · Score: 1

      I went intel+nvidia on the laptop, because core duo is the only high-end mobile computing game in town until AMD makes their next process shrink and catches up. I was tempted to wait for core duo 2 to show up, but I couldn't wait, and got a HPQ nw9440...

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  99. Let me put this in perspective by default+luser · · Score: 1

    So they leave that out of the Free Software driver they release. Big deal.

    You really have no idea how pervasive S3TC is, do you?

    S3TC was released as part of the DirectX standard (now called DXTC). It was a dying gasp from S3 about 6 years ago. It was immediately picked up by card makers (Nvidia has supported it since the GeForce 2, and ATI since the Radeon), and has become an industry standard.

    I guarantee EVERY GAME RELEASED today uses DXTC without even telling you...and you don't even notice it!

    If you want perspective on what that means for users, how about this:

    Doom 3 has an "Ultra Quality" mode, which uses the same resolution textures as the "High Quality" mode, but leaves them uncompressed. The "Ultra Quality" mode requires a SMASHING 512MB ram, and reviewers and players alike can't tell the difference between it and "High Quality."

    The hardware difference? "High Quality" only requires 256MB of ram. THAT is why S3 / DirectX Texture Compression is so critical.

    --

    Man is the animal that laughs.
    And occasionally whores for Karma.

  100. Why would you still think in the old model? by dascandy · · Score: 1

    I recall seeing a post about FPGA's in an Opteron socket, allowing fully customized coprocessors for the Opteron. I also see now that ATI is being bought by AMD.

    Adding 1+1 gives me 3, otherwise equal to, make a GPU that fits an Opteron socket that you can replace just as easily as a normal processor, with its own memory in a similar way to normal outputs and a "videocard" reduced to about a RAMDAC.

    I'm in.

  101. Don't know by agentdunken · · Score: 0

    I don't know if this is a good thing or a bad thing really. This may be a good thing because now ATI may be able to make way better drivers or Windows and Linux. Driver support for both systems have been terrible. Then technology growth may be faster and better since now you have two major power house companies working together.. Just one thing I hope does not die off... The Nforce chipset... I hope we don't loose the Nforce chipset...

    --
    Linux, because a PC is a terrible thing to waste.
  102. Re:Intel has killed gaming...but AMD has restored by Anonymous Coward · · Score: 1, Interesting

    Bullshit, plain and simple.

    As a games programmer who works on AAA PC titles I can tell you that it simple isn't true that the intel chipsets are incapable of running games. You can't treat them as if they're a new generation top of the line GPU but they run games just fine.

    And unlike AMD intel are much friendlier towards helping games companies. Which company runs seminars and training on and off site on multi-core multi-threading for games? Which company sends out development machines to developers so they can test their games on the latest hardware? Which company builds performance and profiling tools and provides source code and samples to developers?

    AMD do nothing for the gamer or the games industry apart from providing some processors that run floating point maths very well however Intel reaaly do make a difference.

  103. Wishing I had moderator points right now... by Svartalf · · Score: 1

    You're 100% correct. And better yet, there's things looming on the horizon that few know about.

    But hey, this IS /., so those proclaimations are dime a dozen. But I strongly suspect that I'll
    have some decent pull now on the driver front (LGP- I'm sure ATI can be convinced to fix things
    so I'm not sitting here twiddling my thumbs on mostly 2D only game ports because the stupid driver
    on my laptop doesn't perform worth spit...) and in 4-7 weeks, things are very likely to get MUCH
    more interesting for all parties involved- and I'll have a bigger impact on these sorts of things.
    (Again, this IS /., so I'm not even going to expect you to believe a single thing I'm typing here
    because talking out one's behind IS the norm here...but I'll re-iterate, things look like they're
    about to change very rapidly.)

    --
    I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
  104. One reason why AMD may have bought ATI by RayDude · · Score: 4, Interesting

    There is one obvious reason for the purchase, already stated by others. I'm just reiterating.

    Next year, AMD will be shipping quad core Athlons and Opterons. But, if they wanted to they could replace one CPU with a GPU and have video on die. And if they wanted to they could replace a second CPU with sound, USB, SATA, Gigabit, wireless etc etc etc, and have an entire computer on a chip.

    VIA has been trying to do this for years. AMD has the fab capacity to pull it off.

    AMD could be the first company to enable the $150.00 PC to exist (by saybe 2009). Smaller than a mac-mini, dual core, and all you need to get it to run is slap some flash memory on board for a hard drive substitute, some DDR2, a cheap DVD drive and Voila! Instant computer.

    Imagine a Dual Core Athy with a gig of ram, 20GB flash disk all in the form factor of about twice the size of an IPOD.

    Oh you could put a screen on it too, DGMS.

    This could be a great thing. My only advice for AMD / ATI is: Dedicate some resources to drivers, or better yet, open source the GPU API.

    Raydude

    1. Re:One reason why AMD may have bought ATI by neovoxx · · Score: 1

      What about a direct HyperTransport link to a socketed video processor?  A link via HyperTransport would boost video performance signifigantly.

      --
      0x68ADA2CC
    2. Re:One reason why AMD may have bought ATI by LesPaul75 · · Score: 1

      Well, what I've read/heard is that the primary limitation to building a true, x86 based system-on-chip isn't the area, it's the I/O. There would just be too many pins to put everything (CPU, GPU, Memory controller, Southbridge, SIO) all on the same chip. So it's not a question of just replacing a couple of CPUs on the die with other chips. But anything's possible, I suppose.

  105. GPUs are too different by BobPaul · · Score: 1

    GPUs tend to be more vector processing units, whereas standard CPUs are less fine tuned for this type of processing. Using a CPU as a GPU would suck.

    Now, AMD has talked in the past about using a 939 (or I suppose AM2 now) socket for either a CPU or a GPU. The motherboard would have the hardware to plug in a monitor etc but a graphics card upgrade would be more of a graphics chip upgrade. Want 2 CPUs? Put in a more normal graphics card into a PCI-E slot.

    They've brought it up a few times; I wouldn't be too surprised to see that happen in the near future.

  106. AMD and nVidia?? by bubkus_jones · · Score: 1

    I wonder what this means (if anything) for those of us with AMD/nVidia systems.

    1. Re:AMD and nVidia?? by justsomebody · · Score: 1

      I wonder what this means (if anything) for those of us with AMD/nVidia systems.

      Few possibilities:
      - NVidia goons show up at your door and forcefully take your chipset
      - Your chipset evolves, grows legs and runs away
      - Your drivers suspiciously start producing BSODs or kernel panic. How the hell did they find out?
      - Nothing, as you already bought your computer and it was well supported?

      --
      Signature Pro version 1.13.2-3 release 83.5 beta3try7 after-breakfast edition
    2. Re:AMD and nVidia?? by bubkus_jones · · Score: 1

      I should have phrased that question better, my appologies.

      What I was trying to get across, was, will this merger/buyout affect AMD's appeal to Nvidia users, or ATI's appeal to Intel users (or even Nvidia's appeal to Intel users)? Should we expect them to start providing "enhancements" that Intel/ATI or AMD/Nvidia users won't get?

      We're already seeing it on a per-game basis, were some games pimp out the "funky-new-feature-of-the-day" that each card is trying to use to lure you to their camp, and the other camp misses out on stuff, but that's all at a software level.

      Maybe AMDTI will start developing a faster bus for their equipment?

  107. That's a really narrow view. by Anonymous Coward · · Score: 0

    I don't use linux, I use openbsd. Obviously I have a windows machine for playing games. I make my hardware purchasing decisions, and the recommendations I make to everyone I know based on what companies are friendly towards open source. Companies that tell me I am meaningless and do not matter don't get my money. This is why I buy amd and not intel. And this could be why I buy ati/amd instead of nvidia, if amd is smart enough to force their friendly, open and helpful attitude onto ati with the aquisition.

  108. Whew! by Brandybuck · · Score: 1

    I'm glad it wasn't ATI who bought AMD, otherwise we would have to use buggy proprietary CPU drivers...

    --
    Don't blame me, I didn't vote for either of them!
  109. The end of traditional CPUs by Anonymous Coward · · Score: 0

    This is just another sign that traditional CPUs have their days numbered, with massively-parallel multi-core designs becoming the norm. Intel probably sees this too, and I wonder what they're doing to adapt.

  110. Re:Hmmmm, Consoles pickup lines by Anonymous Coward · · Score: 0

    Hey, xbox, baby! You look like you got a little AMD in you.

    No, not that I know of.

    Would you like one?

  111. I've decided by Cyno · · Score: 1

    This week I don't hate AMD quite so much.

    See, me and AMD have this love/hate sorta relationship, kinda like SGI. They're one of the good guys, or at least that's been everyone's perception, for the most part. They have so much wonderful technology, but a lot of it is kept just out of the consumer's reach.

    But the quad core challenge is kinda cool, price cuts are welcome (though not quite enough w/o x2 3600+), and this ATI merger has some worried. I think ATI's quality and products have shown considerable improvement. Their Linux support is acceptable, and as long as everything works I dont mind integrated graphics. In fact, I love having a GPU available on a low end system for basic OpenGL when its necessary. Even my laptop's Intel chipset is useable. Anyway, AMD's recent announcements have calmed some of my fears that they've lost focus. I'm still concerned for AMD that it will be easy for me to get a $300 Conroe and overclock it to 2x 3.0 Ghz. AMD has no price/performance comparison to this, even after the cuts, that I am aware of. Maybe they're easy to overclock as well.. I haven't heard much about X2 overclockability, and 3.0 seems a bit high for the 90nm A64.

    I wish AMD would not do things like the 939 Athlon vs. 940 Opteron, cpu clock multiplier locks, disabled chips to prevent multi-CPU functions, and other typical consumer practices that insult my intelligence. But they got rich shareholders to answer to. Shareholders that cost me $100s in new motherboard, memory and peripherial upgrades. Shareholders that probably own shares in motherboard, memory and peripherial manufacturers (Conspiracy, I tell ya!). While I'm at it, I wish Microsoft would GPL Vista, Intel make a HyperTransport Core 2 (for the benchmarks), and someone, for the love of God, please make 10ns Ghz DDR2. thanks

  112. Brain hurt... stop talking already! by Astin · · Score: 1

    I'll try and be short here:

    - AMD needed ATI to compete with Intel's larger product line. Intel ships more graphics cards than both ATI and nVIDIA because of their bundling.
    - AMD has committed to maintaining ATI's entire Graphics line, including Intel products. Although word is that Intel is withdrawing ATI's license to do so
    - ATI also has a core logic business AMD wants
    - ATI has chips in consumer electronics - consoles, HD TVs, etc. AMD wants into these markets as well.
    - Intel makes its own chipsets, and has end-to-end motherboards. nVidia still has chips for these, why wouldn't they have them for AMD as well when AMD starts making their own?
    - Intel and nVidia will NOT be merging. There is too much overlap in their businesses because Intel covers such a large spectrum... which is the reason AMD is buying ATI - to compete.
    - It is NOT official - the offer is official, the acceptance by ATI's board is official, but the actual purchase still needs regulatory approval... which will likely not be a problem.

    This isn't JUST about jamming some low-end ATI GPU on an AMD chip.

    --
    - In hell, treason is the work of angels.
    1. Re:Brain hurt... stop talking already! by Anonymous+Freak · · Score: 1

      Regulatory, (between U.S. and Canadian regulators, since ATI is a Canadian company,) and shareholder (of both companies) approval.

      What I see as interesting is that Intel was probably working to get SLI/CrossFire support in their chipsets. Now will they bother with CrossFire? Yeah, ATI/AMD could write drivers that support it on Intel chipsets, but Intel will probably declare this as 'unsupported' while nVidia's SLI would be officially supported. (Intel knows they still rely on ATI and nVidia for high-end gamer video cards. They don't even attempt to say integrated video is good for hardcore gamers.)

      And, yeah, I can also see them withdrawing ATI/AMD's license to produce compatible chipsets. Which might actually drive more business to AMD, as people choose ATI/AMD combos rather than Intel/Intel combos. (When they would have chosen an ATI/Intel combo before the 'ban'.)

      --
      Another non-functioning site was "uncertainty.microsoft.com."
      The purpose of that site was not known.
  113. The sky is falling... by BlueCoder · · Score: 1

    AMD just destroyed the only competition that nVidia could possibly face. No other graphics company out there has enough IP that they could compete and survive an nVidia lawsuit. It also mean that nVidia no longer has to push it's development and release cycle and can somewhat sit on their ass as they rake in the money. You can't fight two wars on two fronts people. nForce for AMD is dead in two years. Hopefully ATI has enough talent onboard that they can shift from graphics card design and focus on chipset design. The only thing that's going to happen with ATI graphics is it being offered as builtin server video in comparison to intel integrated graphics. The only thing I can imagine is the video team working on slower "workstation graphics" and after five or more years being sold off.

    They obviously have no idea how this is going to taste to all their previous fans. Party at Intels house tonight!

  114. Re:Intel has killed gaming...but AMD has restored by Brigade · · Score: 1

    Whatever .. look at console vs. PC games sales (overall) and comparatively (Oblivion for 360 vs. PC) .. integrated graphics, granted, don't power games nowadays .. but it's a moot point when I can either spend $500 every 6-12 months to get all the bells and whistles out of my PC (and that's JUST the video card) .. or spend $500 on a console that does it cheaper (and with all the extra features on the new generation .. )

    My question is: If the PC gaming market is shrinking, (unless Vista makes for huge demand for new graphics cards) .. shouldn't ATI be considering who they're going to develop for on the next console hardware cycle? I don't see Microsoft abandoning Intel to play nice with AMD/ATI ... and Sony (I believe) has ignored both ATI/Nvidia to go their own way.

  115. Probably not... [Re:Don't really know..] by codergeek42 · · Score: 1

    Hmmm probably not. CPUs are general-purpose processors; whereas GPUs are more specifically designed to do lots of math-intensive graphics work such as texture and vector calculations.

    Now if one of the cores would do something like physics calculations on demand, and the other could do the logistics in a game, then I think it would be VERY cool. 8)

  116. What this means. by Anonymous Coward · · Score: 0

    AMD will have a top of the line graphics chip that supports Coherent Hyper-Transport out the gate.

    You will no longer have to buy video cards for AMD systems, you will buy GPUs and plug them directly in to the motherboard. It will share system RAM with the processor.

    Imagine buying a board with 4(or more) sockets. You start with some basic on-board video technology, nothing fancy. Initially you just have one CPU (dual-core, whatever). Later you realize you miss your favorite 3d games. You go out and buy an A(MD/TI) GPU and drop it in the cHT socket. Voila, instant 3D gaming. It would probably be cheaper, too. No PCB, no extra RAM (though expect the "standard" amount of RAM on systems go to up to 2GB or higher.). A few years later you find the system getting a bit sluggish from the added Vista requirements that came with the newest "Service Pack". Simply slot in another CPU and GPU to "double" your performance.

    This is what this means.

    Once it is out the door and proven with ex-ATI GPUs, Physics co-processors, advanced 3d audio chips, or whatever specialty processor you can think of that will take advantage of this would be only as difficult as dropping the chip in. It gets the benefit of living very close to main memory on a fast BUS, no card overhead.

    Another advantage of this is that we can start moving towards smaller form-factors for mainboards and systems overall.

  117. Due to this.. by paulmer2003 · · Score: 1

    A source reviled to me that Apple is now pulling all of their products off of the shelf. Ya can read more on http://paulmer2003.com/?p=117

  118. So What Do I Do Now? by Stormy+Dragon · · Score: 1

    I'm in the process of building a new gaming computer. I was planning to go with a 975XBX motherboard with a Core 2 Duo processor and ATI Crossfire, but I'm not sure if that's a good idea now. Nothing on Intel or ATI's site says anything about how this deal is going to affect their current Crossfire efforts.

  119. Different kinds of power. by raehl · · Score: 1

    Mass-market CPU's are scalar. GPU's are vector.

    Today's scalar CPU's are NOWHERE CLOSE to being fast enough to do a GPU's worth of vector math.

  120. Re:Intel has killed gaming...but AMD has restored by 0111+1110 · · Score: 1

    Apples and oranges. Console games and PC games are marketed for different demographics. PC games for the old timers and console games for the kiddies and ultra-casual gamers. This is changing somewhat in certain markets where a certain game is ported from one to the other. So you end up with a compromise between the two such that neither the 12 year olds nor the 35 year olds are completely happy or totally frustrated.

    --
    Quite an experience to live in fear, isn't it? That's what it is to be a slave.
  121. I've Been AMD/ATI for years... by PixieDust · · Score: 1

    Been this way for a long time. From a Windows standpoint I've been flying. I'm from from a "User" in Windows, and have never had an issue with an AMD/ATI combo. In fact, I prefer that over anything else. From a Linux standpoint I'm definitely a "User". Don't know my way around too well, but can generally infer thingsw well enough from various man pages, walkthroughs, and a few google searches now and again. While ATI's performance is certainly lackluster under Linux, it's not horrible, and tends to be more stable, at least for me, than nVidia. Perhaps I've the 1 in a billion HW config that works well under both Windows and Linux. *Shrug* Either way, seeing AMD and ATI jumping into bed together is very good. Regardless of whether or not you like the two companies, you have to acknowledge that BOTH of these companies have fostered a great deal of innovation over the years, and working together, will only foster more. There were ATI vs nVidia priuce/innovation wars some years ago, just as there was an ongoing AMD vs Intel war. In the end, it always benefits us, and ultimately, the companies involved (though one invariably has to come out losing eventually, change its tactics, regain some footing, and get right back in). I can't wait to see what happens when this powerhouse really starts ramping things up. I just hope that ATI's cards don't go away. There has been speculation of this helping AMD counter the Centrino type stuff. Admittedly, that'd be great, but when was the last time you saw an Intel Video Card up for sale? Certainly can't find them in retail stores. Hopefully ATI continues as a name, and producer, and simply adds their capabilities to AMD's dev teams, and vice-versa. My thoughts

  122. Re:Intel has killed gaming...but AMD has restored by ichigo+2.0 · · Score: 1

    I can either spend $500 every 6-12 months to get all the bells and whistles out of my PC (and that's JUST the video card) .. or spend $500 on a console that does it cheaper (and with all the extra features on the new generation .. )

    Why do some people always claim this? I have never bought a videocard more expensive than 200e, and have never had any performance problems. I've never seen the point for 500e+ videocards (other than bragging rights for having poor economic sense) because they rarely are much more faster than their medium-end counterparts.

    Sony uses an Nvidia chip in PS3, and Microsoft has already abandoned Intel, it's not an Intel CPU that powers the XB360.

  123. Disturbing thought by Anonymous Coward · · Score: 0

    Ok, what is the chance that tomorrow, we will hear that Intel did buy nVidia? Wouldn't this be great?

  124. I know this ain't gonna be read much... by Black-Six · · Score: 1

    I read on the Fox News website that Nvidia and AMD will continue their CPU GPU cooperation even after the news of the AMD buyout of ATI. Nvidia and AMD are going to use the incorperation of ATI into AMD as leverage to unseat the leading chip manufacturer Intel. This could produce even better graphics performance than previous cooperations. AMD's buyout was to produce cheap, reliable, integrated chipset solutions for entry level users and to gain a better foothold in the market. These two juggernaut chip makers could very well put Intel out of business overnight, imagine that no more cheap crappy Intel chips and nothing but high quality chips at decent prices.

  125. Coming Bad News for Open-Source Drivers by Deviant · · Score: 1

    This fortells a huge shift that is coming in the PC industry and internals. Basically, huge and fast multicore processors will be the only major component of a PC aside from the chipset. It will run mostly software-based drivers for things such as the LAN/WLAN, Sound and Video. The reason we needed daugher cards for these things in the past is that the CPU was not fast enough or well-situated to run all these things in parallel without abysmal performance. But with the advent of multi-core processors one core can do graphics, one can do sound/network and one or two can run your app and it will all just fly. As the movement in the industry is already towards smaller, less components, less power and less heat this centralization was only a matter of time. The CPU manufacturers would love making processor performance truly the be-all and must have of performance again in this way. The video card manufacturers will be wounded with this increased emphasis on the CPU so the CPU manufacturers will merge with them and/or buy them.

    The problem for open-source lies in the fact that a good portion of the actual video "card" product will lie in drivers and that they will not want to share the internals of their design in an open way to their competitors. Everything will be binary and everything will be closed. Linux desperatly needs to have a framework ready in the kernel/OS to accomodate this binary and closed driver reality or the manufacturers will not bother to support much of the community out of a few key distros at best.

  126. Red + Green = Caution by jman.org · · Score: 1

    I'm not entirely convinced that one struggling chip manufacturer purchasing another will really help either.

    I live in Austin, Texas, where AMD has one of their fabs.

    Back around '99 they layed off a bunch of people, some there for 20 years.

    Now they're trying to build a new fab over a very environmentally sensitive portion of town (recharge zone for a huge aquifer that is the primary source of water for folks stretching all the way down to San Antonio. Austin itself gets water from other sources.)

    This new fab has caused a great deal of consternation; some say building is inevitable, so just get over it, some say it doesn't matter how many buildings you have if there's no water for the occupants to drink.

    Personally, after purchasing an AMD 64-bit-based laptop 1.5 years ago, I realized I jumped the gun as there's not enough software in the 'doze world to really make use of the chip. (Alas I must use 'doze for work, so a proper 'nix operating system is not an option for that machine).

    I'd thought I was getting the next batch of technology and 'helping the home team', but with they way they want to trash the environment, my next purchase will definitely be Intel based.

  127. Slashdotters are idiots!! by Anonymous Coward · · Score: 0

    Comments like this, this, and this shows how stupid some Slashdotters are. They made baseless accusations that the analyst was pumping the stock. Now they and those that modded them up know just how stupid they are compared to these analysts. No wonder their jobs are all being sent to India.

  128. Amd+ati by vision864 · · Score: 1

    All the linux crackpots expecting driver extacy aside - does anyone else get the sick feeling that AMD is about to mediaGX themselvs?

  129. We have a Kool-Ade drinker. by LWATCDR · · Score: 1

    Man you are just nuts.

    Your right users don't want the latest version of every program on their system. They want the latest version of the programs they use all the time.
    Yes users do want to go to Com-USA, BestBuy, Circute City, ZipZoomFly, or New Egg and buy the latest printer, digital camera, webcam, scanner, or wifi-fi card and have it just work.
    Good grief. I showed you an example of FOSS that was available in a new version than what was in a distro's repository to show why a vendor written and supported driver might be newer than then on available from the disto. You said "Doesn't matter. People don't need the latest version."
    Then I gave you an example of when a stable binary interface would be an advantage to an end user. Not even theoretical but one that actually happened to me.
    You said, "Doesn't matter no user would want to use the latest hardware."
    What takes the cake is "It sounds like you're saying that because there might be issues with compatibility between versions that it automatically means the community is completely incapable of providing any usable drivers at all?".
    The 3Ware driver is in the kernel. Do you know who paid the programmer that wrote it? 3Ware did. A lot of the people that develop all those FOSS drivers work for the companies that sell the hardware! Even with a FOSS driver you can have lag between what is available, what is available in the Kernel, and what is available in the distro's repository. Any solution that involves using the command line, make, or menuconfig IS not a good solution for the vast majority of the desktop and laptop users. It is even a barley acceptable solution to experienced Unix/Linux admins.

    Man you have almost convinced me that Microsoft is right. With an attitude like yours Linux will never be a good replacement for Windows. Good thing not every developer shares you attitude.

    I will say that the Kernel developers have every right to not create a binary device driver interface. If I really wanted to and had the time I could fork the kernel and put one is. If enough people joined me and supported my version it would be the standard. That is what is great about FOSS. What ticks me off is them not flat out saying why the are not going to put one in and then make up excuses like, security, stability, and speed.

    --
    See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    1. Re:We have a Kool-Ade drinker. by mrchaotica · · Score: 1
      Then I gave you an example of when a stable binary interface would be an advantage to an end user. Not even theoretical but one that actually happened to me. You said, "Doesn't matter no user would want to use the latest hardware."

      Don't go putting words in my mouth! That is NOT what I said!

      First of all, I actually said "average user" (keyword: average). Your example involved configuring a server as part of your job; therefore it was invalid because you are not "average!" Second, I did not say that the average user wouldn't want to use the latest hardware; I said (or at least was trying to say) that the average user would be forced to not even consider that 3Ware card because it was not, in reality, supported!

      Good grief. I showed you an example of FOSS that was available in a new version than what was in a distro's repository to show why a vendor written and supported driver might be newer than then on available from the disto. You said "Doesn't matter. People don't need the latest version."

      And here you made the same misconception: I wasn't trying to say that "people don't need the latest version," I was trying to make the point that, for an average user, the distro's version is the latest version!

      Here's the underlying point I'm trying to make: "average users" don't use Linux. They use "Fedora Core" or "Ubuntu" or whatever other distribution. There's no such thing as "supporting Linux," because as an entire operating system, it doesn't exist!

      An average user won't install a driver because -- even if some other version of Linux supports it -- any hardware not supported by the distro's stock kernel is not supported at all. An average user won't install software outside of the distro's repository because, for him, it's not the "latest version", it's an unsupported beta version. For the distro user, it doesn't exist because it hasn't been released into the repository yet!

      Now do you understand what I'm saying?!

      Any solution that involves using the command line, make, or menuconfig IS not a good solution for the vast majority of the desktop and laptop users.

      Darn right! However, a separate binary driver is not a good solution for them either. The only good solution is -- going back to the original topic -- for Free Software developers to have the specs so that they can make drivers capable of going through the testing and verification process that's required for inclusion with distros' kernels.

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    2. Re:We have a Kool-Ade drinker. by LWATCDR · · Score: 1

      And what I am saying is that having to depend on the kernel developers and distros isn't practical. What I am saying is average users want to go to Best Buy and buy a scanner, printer, or web-cam that supports Linux. The problem is that without a stable binary interface the vendors can not put a driver on a CD or a web and tell the user just install this or download this and have it work.
      Average users on windows install drivers all the time and it works just fine. Right now there is NO way for a hardware manufacture can say supports Linux because they can not know if the distro includes the driver. Not only that they can not possibly me sure that Linux will support it when it is shipped because than can not wait until the Kernel developers to include the driver in the kernel and then wait for it to trickle down to the distros.
      Until there is a binary driver interface Linux will always lag in hardware support of will be stuck with complex installation procedures like the one for the nVidia driver.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    3. Re:We have a Kool-Ade drinker. by mrchaotica · · Score: 1
      And what I am saying is that having to depend on the kernel developers and distros isn't practical.

      On the contrary, it's the only method that is practical! It works for Mac OS, you know. The fact that Apple maintains basically all the drivers (including external hardware -- not once have I ever actually used any driver that came with the hardware with my Mac; I've always just plugged the thing in and let Mac OS handle it) is a big part of why it has the reputation of being much more stable than Windows.

      Hell, even Windows is moving to this model, to an extent: in order to be installed without a warning, Windows drivers have to be certified by Microsoft. If I recall, sooner or later they're going to start requiring certification in order to install. Granted, the drivers aren't actually being developed by Microsoft itself, but they're still checking it -- in exactly the same way the Linux kernel devs or distro maintaniners would for a Linux driver!

      Right now there is NO way for a hardware manufacture can say supports Linux

      No shit, Sherlock! That's why I've been saying that the very idea of "supporting Linux" makes no sense! That's why companies that make commercial software for Linux (e.g. Oracle) don't support "Linux", but support "Red Hat Linux" or "SuSe Linux" instead.

      Until there is a binary driver interface Linux will always lag in hardware support of will be stuck with complex installation procedures like the one for the nVidia driver.

      And if a binary driver interface for Linux ever does happen, Linux will be stuck with shitty, unstable, insecure drivers just like Windows. Even the current situation is better than that!

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    4. Re:We have a Kool-Ade drinker. by LWATCDR · · Score: 1

      "That's why companies that make commercial software for Linux (e.g. Oracle) don't support "Linux", but support "Red Hat Linux" or "SuSe Linux" instead."
      And you don't see this as a problem? You think this is a good thing?

      "And if a binary driver interface for Linux ever does happen, Linux will be stuck with shitty, unstable, insecure drivers just like Windows. Even the current situation is better than that!"
      How is that? We do have binary drivers. ATI and nVidia are two very common ones. The Mac used binary drivers, is it stuck with shitty, unstable, and insecure drivers?
      Besides I am not arguing that closed source drivers are better than FOSS drivers. I am saying that a binary interface would benefit FOSS drivers as well as closed source drivers. It would make it easy to provide a binary package just like just about every FOSS application does but in this case for drivers.
      Yes it would benefit closed source drivers as well but I wouldn't mind it they required any driver using them to be FOSS. How you would do that, I don't know.

      Of course the real solution to insecure and unstable drivers is to move them into user space and out of the kernel.
      I know that would cause a performance hit but with modern hardware it might be acceptable for the vast majority of users.
      A way around that problem would a tiered driver system. Trusted FOSS driver could reside in kernel space while the less trusted binary drivers live in user space.
      This would have some advantages in that even FOSS drivers could start off running in user space for testing and once they have proven themselves stable and secure they could then be migrated into kernel space. Or if the device was had low performance needs like a keyboard or web cam it could stay in user space for the extra stability and security that would offer. To me it seems an unnecessary risk to allow something like a bug in a serial port driver to cause a kernel panic. I can see where some device drivers need all the performance that they can get. RAID controllers, network cards, and graphics cards all jump to mind but things like web cams, printer ports, and serial ports do not. I would seem logical to offer a variety of performance to security trade offs as far as device drivers go. I know this is a different subject but if you are worried about security there are ways to improve that and add a stable binary interface.

      As computers get faster and faster I am beginning to think that Hurd and Minix3 might be the way to go. It may now be time to trade some performance for reliability and security. Modern CPUs might make microkernals the way to go. Of course I hope if that becomes the case that Linux will migrate that way.
      Also I really find it strange that you think that closed source drivers would be shitty, unstable and insecure while FOSS are all perfect. Guess what there are FOSS drivers that suck just as bad as closed source drivers. A good number of drivers in the kernel are not maintained as well as they should be. The person that was maintaining them has lost interest and no one has stepped up to replace them. I wonder how many krufty old drivers are sitting in the kernel source that have not been looked at in years. Luckily most of those are not compiled into average distros kernel so they rarely cause problems. Just as Microsoft is wrong thinking that FOSS can never be as secure and bug free as Closed Source software you are wrong in thinking that every piece of FOSS is secure and bug free. Take a look back at the problems in some of the release versions of 2.6 before the switch in VM systems. FOSS is great but not perfect.
      I can honestly tell you that I have not had any issues with drivers on my XP machine. I have seen problems with drivers under Windows but then I have seen them under Linux as well. You example of the Mac is another good one of good secure closed source drivers.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    5. Re:We have a Kool-Ade drinker. by mrchaotica · · Score: 1
      That's why companies that make commercial software for Linux (e.g. Oracle) don't support "Linux", but support "Red Hat Linux" or "SuSe Linux" instead.
      And you don't see this as a problem? You think this is a good thing?

      No, actually I think everything ought to be compiled from source, so that it can all work together (e.g. Gentoo, broken Portage issues aside). But if you're going to insist on distributing binaries, you have to know what all the other library versions, paths, etc. are on the target platform in order to not have everything all screwed up. The only practical way to achieve that is to have the whole platform controlled by a single vendor (e.g. Microsoft, Apple, Red Hat, Debian).

      A way around that problem would a tiered driver system. Trusted FOSS driver could reside in kernel space while the less trusted binary drivers live in user space.

      This is an interesting idea, and it might work fairly well. I'm not sure I like it, though, because in addition to still being suboptimal compared to having all Free drivers, it also seems similar to stuff Microsoft is trying to do with Treacherous Computing. I fear that if this kind of idea took off Microsoft's model (which is just the opposite: the binary drivers are "Trusted" and the Free ones aren't, and only "Trusted" drivers can run in kernel space) would gain a lot of traction too because most people wouldn't be technically inclined enough to understand the difference.

      Hopefully we can at least agree that we don't want to go the Microsoft route (where the kernel will be able to act against the user because it will have access to secret encryption keys and will disallow learning those keys by refusing to run un-"Trusted" code in kernel space), right?

      Also I really find it strange that you think that closed source drivers would be shitty, unstable and insecure while FOSS are all perfect.

      It's not that I think closed-source automatically implies low quality; it's that closed-source automatically makes it extremely difficult to determine the quality. With Free drivers they may be good or bad, but there's a certain peace of mind that comes from knowing that, for example, it's relatively hard to hide rootkits and such in them because the code is there for all the world to see. A closed driver, on the other hand, could be doing anything and you just wouldn't know (practically speaking).

      --

      "[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz

    6. Re:We have a Kool-Ade drinker. by LWATCDR · · Score: 1

      "This is an interesting idea, and it might work fairly well. I'm not sure I like it, though, because in addition to still being suboptimal compared to having all Free drivers,"

      Again you are using choosing religion over usability and technology and in this case stability and security.

      I used binary drivers VS kernel mode drivers. That is not a thing in the world that would prevent a company ore a group from releasing a FOSS binary drivers. The source would be available but an easy to install binary would be available as well. A side effect would be that binary only drivers would also be available and easy to install instead of the case we have now of binary drivers being available and a pain to install.

      The other benefit would be that none performance intensive drivers like I2C, serial ports, and other low performance drivers could be moved out of the kernel space and into user space. Also new drivers can start off in user space and then be moved into kernel space after they have been tested and have a track record. Like I said it is not the best of plans when a bug in a serial port device driver can cause a kernal panic. And some users may choose to run drivers in user space that are available in the kernel. In that case the admin could dertermin the trade off in security vs speed. Have a network card that you are not to sure about. Maybe there have been some bug reports but nothing found yet? Take it out of kernel space and run it in user space. If the driver crashes you would have a lot more debug information than if it took out the kernel and you would even have an option for the kernel to restart the driver.
      Server stays up and you get more information to track down the problem.
      And Actually it is the exact opposite of the Microsoft's trusted drivers. The user gets to choose what drivers are trusted and not. I would also like to see a stable binary trusted interface so I could choose to user a binary only driver in trusted mode but I know that the FOSS faithful don't want to give the user the option. This seems to me the better compromise than what we have now. Closed drivers would be easier to install then they are now and less of a threat to the kernel's stability. Open source drivers would still function better and FOSS drivers now have the option to run in user space if that would be a good trade off for them.
      As I said before WE DO HAVE BINARY ONLY DRIVERs. nVidia and ATI just release a source code stub that talks to the binary driver. All the current system does is make every-body's life harder. User and developer alike.

      "No, actually I think everything ought to be compiled from source, so that it can all work together (e.g. Gentoo, broken Portage issues aside). "
      Again not very user friendly for mom to have to spend 30 hours compiling OpenOffice. And it isn't just Portage trust me. I have had more problems trying to compile tar balls than I can shake a stick at. Everything from changes in GCC between 3 and 4 to hunting down header files.
      Not as many problems since I have moved to Open Suse 10.1 but still a lot more than I want.

      I also strongly disagree with the idea that average users don't want the latest version of programs. I do keep the programs I use every day up to date. I work for a software company and I can tell you that when a new version of our software goes up on the website a good 60% of our customer base download it within three weeks.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.