Slashdot Mirror

Search the following pdf for juicy phrases like "he had a gun to his head" and the word "threatened" to learn more about why vendors might allegedly be frightened to deliver AMD-based products.

http://www.amd.com/us-en/assets/content_type/Downl oadableAssets/AMD-Intel_Full_Complaint.pdf

You can view the PDF as HTML at:

http://64.233.161.104/search?q=cache:BK_UrlmGznsJ: www.amd.com/us-en/assets/content_type/Downloadable Assets/AMD-Intel_Full_Complaint.pdf+amd+intel&hl=e n&lr=lang_en

And let me guess, you think lines of code is a good indication of how good a program is, right? The lower the number of transistors, the better; less transistors means less heat, less power usage, and less complexity. It is merely an unfortunately fact of life that adding more functionality tends to require more transistors.

While we're on numbers; why is 100W good? What sort of power consumption do the G5 chips have? For a comparison, AMD Athlon 64 chips have a power consumption of 67 to 89W, depending on the exact chip.

This is not a matter of writing poor compilers. That's not what's being discussed. When you have ~90% market share and have bought every OEM you simply can't by law do what Intel is alleged to have done which is case of corporate sabotage. Again, sabotage is unlawful.

I've managed to read up to page 17 of:

http://www.amd.com/us-en/assets/content_type/Downl oadableAssets/AMD-Intel_Full_Complaint.pdf

Now, for me, page 15 is where things begin to get "juicy". I would think that by now AMD would find all the AMD-friendly desktop/laptop computer makers it can and set up AMD outlets, a la Apple stores. With the profit margins so low on computers, and given that Intel is pushing AMD to the brink of "do or die" or at least putting oil on an already black-iced road for AMD, could it hurt AMD to just go "all out" and relentlessly push it's CPU's? I would hope or suggest they tap into Open Source developers and maybe take two courses of action:

1. De-geek many Open Source user applications 2. Push AMD and Linux hardware and computer arrangements.

After all, IFFF they are deeply in bed with msoft, then it's only a matter of time for Intel to assert a certain amount of "hypocrisy" on the part of AMD, when ms is a convicted and morally defunct monopolist. For any company to be in bed with ms when that company's intractable enemy also in bed with msoft probably spells the reality or likelihood of pissing off ms.

According to this posting http://devforums.amd.com/index.php?showtopic=34&hl =intel+compiler on the amd developers forum it seems to have started with version 7.0.

Also amd gives many hints on optimizing for amd cpus with the intel compiler in there Compiler flags document (PDF)
http://www.amd.com/us-en/assets/content_type/white _papers_and_tech_docs/32035.pdf
It mentions that some specific optimization flags work for version 7.1 and 8.0 but not for version 8.1.
It seems that intel improves its amd-specific optimization with every version. :-)

I could not find any other postings on this topic on the amd forum. You would expect people to post these kind of issues on such an obvious place.

Wimp

"Premature optimization is the root of all evil" Donald Knuth

According to this posting http://devforums.amd.com/index.php?showtopic=34&hl =intel+compiler on the amd developers forum it seems to have started with version 7.0.

Also amd gives many hints on optimizing for amd cpus with the intel compiler in there Compiler flags document (PDF)
http://www.amd.com/us-en/assets/content_type/white _papers_and_tech_docs/32035.pdf
It mentions that some specific optimization flags work for version 7.1 and 8.0 but not for version 8.1.
It seems that intel improves its amd-specific optimization with every version. :-)

I could not find any other postings on this topic on the amd forum. You would expect people to post these kind of issues on such an obvious place.

Wimp

"Premature optimization is the root of all evil" Donald Knuth

Sure, they are a corporation and they want profits. The problem is, AMD tried lowering prices to gain share and it led them to nowhere fast. Intel is forcing AMD into a corner, by making abusive deals with OEMs.

Even giving one million CPUs for free to HP did not get them a toehold in the business desktop segment.

AMD did make a performance tool called Code Analysthttp://www.amd.com/us-en/Processors/Develop WithAMD/0,,30_2252_3604,00.html. From the little I have used it, it works great. They also have some good guides on how to improve your performance. AMD isn't up to the same level as Intel with there Developer support, but they have provided some great tools and white papers to help with performance.

While those may be the 8.0 tests as opposed to the 8.0.1 tests, it strikes me that the testing on linuxhardware looks a bit funny. The benchmarks on AMD's site are for the opteron 150 and the piv 3.4 ghz (w/ 1 mb of l2 cache). The ratings are about neck and neck on the amd site but about twice the speed as on linuxhardware's site.

The actual piv that linuxhardware test actually (model 670, piv) has 2 mb of l2 cache and clocks in at 3.8 ghz and for some reason is slower than what AMD got on for a slower chip?

This may be a compiler issue, which at the end of the day says benchmarks are meaningless until you use the right compilers

before anyone responds to this by saying well they used the same compiler so it is a fair benchmark, it is not. That benchmark tells you how long the compiler people spent optimizing for a particular chip in contrast to another chip.

No, Intel bashing is perfectly justified. Here is a comprehensive list of reasons why you should hate Intel:
http://www.amd.com/us-en/assets/content_type/Downl oadableAssets/AMD-Intel_Full_Complaint.pdf

contains some amazing revelations

And processors basically do operate flawlessly, because we demand that from them. But Operating Systems and other pieces of software do not.

You'd better tell AMD that they can take their errata document for the Opteron down then.

Software and hardware is "bad" because noone is willing to pay for "good", NASA's JPL make software that is as close to perfect as anyone can do ... and they spend a huge amount of money to do so.

Actually AMD has a really nice chip that competes with the Xscale http://www.amd.com/us-en/ConnectivitySolutions/Pro ductInformation/0,,50_2330_6625_12409%5E12410,00.h tml
IBM and Freescale also have some PPC chips that are used in embedded systems that could have also worked for the IPod.
The Dell comment does make me think though. I would if it not the server market more than the IPod that is driving the change.
It is very likely that IBM is limiting Apples access to server cpus. Why are there now 4 or 8 cpu Apple servers? Maybe IBM does not want Apple to compete with IBMs Power based servers?
Intel would have no problem with selling Apple any thing they want.

The AMD Turion is already up to 2.2Ghz and down to 25W...oh, and unlike Intels offerings, it is already 64bit capable.

I'll agree as soon as I'm allowed to import products from there for the same prices they're paying, then I could work for the same wage. At the moment companies are allowed to have the work done elsewhere, but at the same time you can't buy the cheaper products there and sell them over here.

Last I checked, prices in Germany weren't that much different than prices here in the US. Some things are even more expensive.

As far as unsavory tactics are concerned: not all litigation is bad litigation. There are times when it is fully justified, particularly in the case of an anti-trust suit. Such things arn't thrown around lightly. They're hard to prove & thus quite expensive.

"Please, AMD do not use this in some SCO-like attempt to pump your stock price. Instead, advertise your products. Let the courts decide whether or not Intel has had an unfair advantage. Mud slinging just makes you look like you are hiding something."

If you read the complaint your head would probably spin with some of the allegations against Intel. Granted, the complaint is only AMD's side of the story, but if even half of the allegations are proven, Intel deserves a serious spanking.

After reading the whole AMD filling, I'm starting to doubt if the Tyan Tiger S2875 and the S2875S aren't dual core ready because of a/an Intel "quota" requirement... if so, i'm thoroughly pissed off at intel now. Looking at Tyna's lineup, there are other opteron boards not dual core compatible, and just as many that are according to this: http://www.tyan.com/support/html/cpu_athlon_duron_ opteron.html I can't seem to find a particular pattern that might say the S2875 should be dual core ready, and I don't know if Tyan is a staunch AMD supporter, it seems like it is (it also seems like anyone would have to be to put up with Intel and AMD at the same time, from the full brief I read). This is all I can find other than emailing Tyan myself: http://forums.amd.com/lofiversion/index.php/t45729 .html BTW, I have the Tyan Tiger K8W S2875

Since it shares manufacturing/fabrication capability with IBM, it has run into many of the same manufacturing and supply problems as IBM.

Since it shares manufacturing/fabrication capability with IBM, it has run into many of the same manufacturing and supply problems as IBM.

Read points 33 and 34 from TFC

You'd think if Fujitsu, HP, IBM and Acer were having such a hard time eking out a profit with Intel processorsd, they'd be MORE likely to try things out with AMD, no?

I'm not saying that AMD doesn't make some valid arguments here, but I fail to see how Intel having their OEMs struggle to make a profit forces those OEMs closer to its hand when there's a viable alternative?

Site claims to support RealPlay, but not for this event:
http://www.amd.com/breakfreewebcast

They say they will host the audio for ten days, maybe they will re-encode it after the presentation.

You might want to read the complaint before you come to such an abrupt, erroneous decision.

Unless, of course, you're just astroturfing.

Full text of the complaint filed can be found here in PDF format.

Interesting read...it's high time we saw some legal action against Intel for all these shenanigans. However, I'm doubtful that this will resolve anything...in reality, Intel will probably be about as inconvinenced by this antitrust action as Microsoft was by theirs.

The source is always the best place to go for information. AMD's page, above, has various links and pages of information on how and why they're creating this new plant and what the details of its operation are.

The source is always the best place to go for information. AMD's page, above, has various links and pages of information on how and why they're creating this new plant and what the details of its operation are.

AMD Fab 36 is expected to be in volume production in 2006. The new facility is expected to require approximately 1,000 employees, most of them highly skilled engineers and technicians.

Source: http://www.amd.com/us-en/0,,3715_10023,00.html

AMD has got a webcast of the event, and the presentations in PDF format here: http://www.amd.com/us-en/Corporate/InvestorRelatio ns/0,,51_306_10383,00.html

"As we look ahead, although we've got some great products now, we can envision some amazing products we want to build for you. And we don't know how we can build them with the future PowerPC road map," Jobs said during his keynote.

The problem with the future PowerPC chips is performance per watt, Jobs said. Intel's chips are far ahead of IBM's when it comes to delivering performance without consuming a lot of power, a quality that is very important to Apple's future products, he said."
[end excerpt]

Jobs is looking for better "performance per watt" and picks Intel over AMD which was not a very smart decision on his part. Apparently he is unfamiliar with the newest AMD 'venice' core and the derivative 'Turion' AMD mobile chips which offer better performance than the Pentium M with less power consumption.

These tables seem to confirm that. Look at the Winchesters on page 13. The 3000 runs at 1.8GHz with a TDP of 67W. The 3500 (2.2GHz) can be run at 1.8GHz with a TDP of 46W.

Has this ever been independently and scientifically confirmed? I would be very interested in the results.

The implication being that if the 3200 is working as hard as the 4000 is, the 4000 should be able to run slower with less work.

These tables seem to confirm that. Look at the Winchesters on page 13. The 3000 runs at 1.8GHz with a TDP of 67W. The 3500 (2.2GHz) can be run at 1.8GHz with a TDP of 46W.

Has this ever been independently and scientifically confirmed? I would be very interested in the results.

To see ADA3500DAA4BP from http://www.amd.com/us-en/assets/content_type/Downl oadableAssets/K8_Electrical_Spec_Rev_ENG.pdf

I'm programming "general purpose x86-64 + SSE3 only" on Linux-2.6.12 & FreeBSD-5.4.

A64 3500+ Venice - 4 GiB DDR400 - Asus A8V - 250 GB HD RAID1.
-=-=- DaNgErOuS hAcKeR -=-=-

published numbers seem to vary a LOT
Interesting. I haven't noticed that myself.

and I'd love to see some hard data on the min, mean, and max draw of the 90nm Athlon 64s

Well you can start here.

As you can see from this white paper. AMD indicates that a 3500+ (Winchester core) has a TDP of 67W at 2200Mhz/1.4V, 56W at 2000/1.35, 46W at 1800/1.3, and 20W at 1000/1.1. Is there some kind of mystery here that I am missing? The numbers have been tested and published by AMD themselves (who would not exactly be inclined to post falsely high numbers).

only eats between 7 and 35W

Huh? Do you have a reference?

Pentium III line came in at the low 30s) with a theoretical max somewhere in the 60W range.

The TDP for a Pentium III-S is 29.97W even at 1.4 Ghz. Yes that is the maximum power. I realize you may have been referring to the older Coppermine core and not the newer Tualatin core. The numbers for the Coppermine core varied a lot depending on clock speed.

Isn't this what CPU Frequency Scaling is for? I personally use the ondemand scaling governor made available in linux kernel 2.6.10, I believe.

Is it really a surprise that the proc will run stably underclocked?

They've been doing this at the factory for some time now.

People would be running OS X on AUD$399 white-boxes instead of buying AUD$799 Mac Minis. They'd lose their hardware markup and a lot of control.

OTOH if people could "run Mac on their PC", Apple might just make money as a software company.

OTGH, maybe "x86" means AMD and being able to plug up to 8 low-power HyperChannel-connected 64-bit CPUs into your laptop. Well? I can dream if I like. (-:

To see ADA3500DAA4BP from http://www.amd.com/us-en/assets/content_type/Downl oadableAssets/K8_Electrical_Spec_Rev_ENG.pdf

I'm programming "general purpose x86-64 + SSE3 only" in Linux-2.6.12 & FreeBSD-5.4.

I've VMware running W2K SP3 as guest.

A64 3500+ Venice - 4 GiB DDR400 - Asus A8V - 250 GB HD RAID1.
-=-=- DaNgErOuS hAcKeR -=-=-

are opterons are super super fast and AMD kindly, and without NDAs, provides technical documentation on them. that's why i buy them

In addition to the aforementioned Intel chips there are also AMD MIPS based solutions that you may look into.
The Au1550 Security Engine runs seperately from the processor (albeit using the same memory) and has a RNG built in. The RNG uses entropy by letting enough noise build up in a pair of ring oscillators and then querying from them.
The throughput for the RNG is also fairly high and you can get 412.5K words per second.

More info.
http://www.amd.com/us-en/ConnectivitySolutions/Pro ductInformation/0,,50_2330_6625_10509,00.html

...because your computer does not belong to you anymore. It is Microsoft's property now.

Unless you install GNU/Linux!

Oh wait... but an Intel rep confirmed the 945 would help implement Microsoft's DRM at a chip level! AMD all the way!

http://www.amd.com/us-en/Processors/ProductInforma tion/0,,30_118_8796_9240,00.html
They have 30W CPUs at 1.4GHZ, or 55W at 2GHZ. Pitty they fucked up the advertising for them so badly that none were ever sold.

AMD manufactures 100% of its production cores in its facility in Dresden, Germany.

Cool N Quiet downclocks when idle and ramps back up when you need it.

pdf with description of the changes and if that doesn't work
http://www.amd.com/us-en/Processors/TechnicalResou rces/0,,30_182_739_7044,00.html

pdf with description of the changes and if that doesn't work
http://www.amd.com/us-en/Processors/TechnicalResou rces/0,,30_182_739_7044,00.html

I've always built my own systems from components. Doing so may cost more than a pre-built unit, but the big advantage to DIY is that you get to pick every single component based on YOUR standards. You're not constrained by a group decision by some collection of profit-hungry BoD members.

When building any system, it is CRITICAL to match what you want in a computer to the components chosen, and there is a definite trade-off between system speed and stability.

Example: I have no interest whatsoever in 99.9% of the available computer games, but I also do some high-end EDA work (notably electronic schematics, circuit simulation, and PC board design). This calls for computing power that is at least somewhat comparable to systems built specifically for gaming. I also knew that stability (as in resistance to any sort of crashing, application or system) was critical to me.

Like you, I don't like a ton of extra stuff on my motherboards. With that in mind, I chose a Tyan dual-CPU board for the AMD Athlon MP chips, specifically the Tiger MPX. I kitted it out with two processors right from the get-go (Athlon MP2600's), and a full gig of ECC DRAM. The board itself comes with two serial, one parallel, two USB, and a 3Com 10/100 NIC all on board. Everything else is left to the end user. These days, that's about as bare-bones as I've seen any board get.

Tyan motherboards have a great reputation for stability in the face of all kinds of different configurations, but they're not very overclockable and, thus, not very popular with the gaming crowd. Tyan is the board you choose if you're building a serious server or high-end workstation, not something to game with.

One other thing that is absolutely vital if you're serious about DIY: Follow the CPU manufacturer's recommendations for motherboards, power supply, memory, and cooling. I cannot stress this strongly enough, particularly where the motherboard and power supply are concerned! Besides the Tyan mo-bo, I opted for Corsair memory DIMMs and a 550W power supply from PC Power & Cooling, all in accordance with AMD's recommendations for the MP series CPUs. A visit to their site will provide you with hardware recommendations for any of their CPUs, and it looks like Intel provides a very similar aid on their site.

The system has been with me for nearly two years now, and I've had ZERO trouble with it. Windows 2000 Pro (you won't catch me using XP, ever) has been solid as a rock on it, as have every single one of my applications.

Building your own system has a lot to say for it but, as others have pointed out, be prepared to pay more than you would for a pre-built box.

Happy tweaking.

Interestingly enough, AMD seems to be hiring in the UK: http://www.amd.com/us-en/Corporate/Careers/0,,51_8 2,00.html

You must be kidding me. newegg currently only lists 2 Venice core A64s: 3200+(2.0GHz) at $209 and 3500+ (2.2GHz) at $315. Where did you come up with the 4000+ at $643?
AMD's own listed price for bulk here is only $482 for 4000+, which will likely keep it quite a bit below $643 (for comparison, 3500+ is $272 in bulk)

I think the motherboard I bought needs its BIOS upgraded, and I don't have any Athlon64s on hand. If it can't POST, I can't upgrade the BIOS.

That's not always true. This page gives instructions for an Asus A8V (yes, you'll need a floppy drive, but they haven't changed in 15 years so you can just pull one from an old computer). Other boards probably have similar features, check the manual.

The interesting part of such a rumor could be AMD who yes make "Intel-Compatiable" chips ,don't they use a POWER core with hardware translation layer to support x86 instructions?

No, they don't, unless RISC86 is POWER, which I have no reason to believe it is (AMD says the RISC86 instructions were "specifically designed with direct support for the x86 architecture while obeying RISC performance principles"), and they still use that scheme in their current processors. Seee the Nx568 product brief.

AMD is German!

Um, no.

AMD is and always has been a U.S. corporation, headquartered in Sunnyvale, CA from the very first year of its existence. They do have a fab in Dresden, but that makes them German about as much as Nike's sweatshops make them Chinese.