Slashdot Mirror


Big-Iron to Open Up for AMD

vincecate writes "Traditionally the key chips that have allowed companies to scale multiprocessors to large numbers have been proprietary. Some examples are the Cray SeaStar, SGI NUMAlink, HP sx1000, and the IBM X3/Hurricane. This proprietary paradigm is about to change to a more open one. Two companies have developed key chips for building large Opteron multiprocessors, and they will be commercial off-the-shelf parts. PathScale has released InfiniPath which can be used with an Infiniband switch to make a high-bandwidth low-latency interconnect for a supercomputer cluster. The other company is Newisys, which will soon release the Horus chip. This chip will make it possible to build 32 socket (64-core) shared memory Opteron systems."

179 comments

  1. it's about time by qwertphobia · · Score: 4, Funny

    It's about time! That z990 under my desk just isn't fast enough :-)

    --
    Never ask for directions from a two-headed tourist! -Big Bird
    1. Re:it's about time by csirac · · Score: 3, Funny

      You must have a tall desk :-)

    2. Re:it's about time by Anonymous Coward · · Score: 0

      Then get a new z9 it will really keep your toes warm

  2. Expect to see.... by 8127972 · · Score: 4, Funny

    .... an Alienware game system with this chipset by the end of the week.

    --
    This is my opinion. To make sure you don't steal it, it's covered by the DMCA.
    1. Re:Expect to see.... by Anonymous Coward · · Score: 1, Funny

      I'm holding out for the laptop

    2. Re:Expect to see.... by Hal_Porter · · Score: 2, Informative

      Actually, I don't think it would help much. Most games now don't benefit from 2 way SMP, so the benefit from 64 way is debateable to say the least. Still for servers, this thing might help. I suspect that most server applications/os's will have servere scaleability problems once you go this far SMP though.

      BTW, Has anyone heard of the MLX1. Makes you wonder what would happen if you put a bunch of these on a chip with some clever caching and the mother of all memory controllers. x86 Niagra anyone.

      --
      echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
    3. Re:Expect to see.... by Anonymous Coward · · Score: 0

      I run a Battlefield 2 Server on a 4 way (dual dual-core) Athlon-64-X2 box - and the load is high on all CPUs. Though the game is mostly network bound, the server definitely benefits from all 4 CPUs.

    4. Re:Expect to see.... by Anonymous Coward · · Score: 0

      You better weigh about 1000 pounds for that 32 CPU laptop to fit on your lap.

      Oh wait. This is /. Nevermind. You'll be OK ;)

    5. Re:Expect to see.... by nothingbutcoupons · · Score: 1, Funny

      With the natural progression of names for AMD processors (Athlons, Semprons, Opterons), wouldn't it be cool if AMD could get the rights and the next kick-ass Alienware systems were made from Klingons?

      --
      Nothing But Coupons - Your no-frills site for online coupons and discou
    6. Re:Expect to see.... by Anonymous Coward · · Score: 0

      Well I'm waiting to see if anyone can get Linux to run on it...

    7. Re:Expect to see.... by Peter+La+Casse · · Score: 3, Funny
      Actually, I don't think it would help much. Most games now don't benefit from 2 way SMP, so the benefit from 64 way is debateable to say the least.

      64 processors will let you run a lot more spyware before your frame rate is affected.

    8. Re:Expect to see.... by TemporalBeing · · Score: 1

      I suspect that most server applications/os's will have servere scaleability problems once you go this far SMP though.

      Well, if you're running Microsoft OS's yes. If you're running Unix or Linux - that's a different story. I remember seeing results for up to 32-way & 64-way testing by ODSL of the 2.6.0 kernel before it was released. The OSDL site of 2.6 Results doesn't seem to list them any longer - only up to 8 way. The results seemed to show Linux handled 32-way quite well, but were issues going to 64-way at the time. Don't know if they fixed that though - but I would gander so if they were testing for it.

      --
      Truth is like the sun. You can shut it out for a time, but it ain't goin' away. - Elvis Presley (source: imdb.com)
    9. Re:Expect to see.... by Anonymous Coward · · Score: 0

      Most games now don't benefit from 2 way SMP

      True? Or are you just guessing again?

      I think you're guessing.

      At a minimum, single-threaded games running on an SMP system can hog one of the cores while the other core handles the O/S and other tasks. Probably not much of a speedup, but given the lengths that players go to in order to get +3% framerates...

      My next game box will be a dual-core. After working with a dual-CPU machine, I'm really looking forward to upgrading from a single-core on the game box.

    10. Re:Expect to see.... by Cromac · · Score: 2, Funny
      Actually, I don't think it would help much. Most games now don't benefit from 2 way SMP, so the benefit from 64 way is debateable to say the least. Still for servers, this thing might help. I suspect that most server applications/os's will have servere scaleability problems once you go this far SMP though.

      Fortunately all computing isn't about games. With this setup you'll be able to encode those DVD rips to DIVX in seconds instead of hours!

    11. Re:Expect to see.... by Hal_Porter · · Score: 2, Interesting

      I know it's a non-traditional here, but I looked up the benchmarks when the dual core chips came out.

      E.g.

      http://www.anandtech.com/printarticle.aspx?i=2410

      "Gaming performance is, currently, highly based on single-threaded performance and thus, we see no benefit from dual core. The thing to keep in mind here is that AMD's dual core solutions are closer to their fastest single core offerings in clock speed, so they end up performing more like their Athlon 64 counterparts in games - which has always been quite strong."

      Look at Doom III, the fastest dual core chip runs slower than the fastest single core. In fact you can pretty much predict performance from clock speed, regardless of the number of cores. So you'd be better off getting the fastest single core chip if that's what you care about.

      --
      echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
    12. Re:Expect to see.... by dwater · · Score: 1

      Actually, *real* games, such as commercial flight simulators and planetariums, do benefit nicely from multiple processors. Many will use libraries that can do this almost automatically - SGI's OpenGL Performer for example.

      Even some big entertainment games, like DisneyQuest, use OpenGL Performer to take advantage of the many CPUs.

      --
      Max.
  3. Do not count out Sun by Anonymous Coward · · Score: 2, Insightful

    Give it another few months and I'm sure Sun will have some server with an obscene number of opterons in it, if thier current direction is any indication ....

    -GenTimJS

    1. Re:Do not count out Sun by Anonymous Coward · · Score: 0

      I could swear that 200-CPU supercomputer I use everyday is a) built by Sun and b) uses Opterons. Wait - it does!

      I don't know what parallelism paridigm it implements though, as I just run stuff on it.

    2. Re:Do not count out Sun by Somegeek · · Score: 1

      But the Sun solution is a proprietary solution. This article is specifically about chips that will allow NON-proprietary solutions that Dell or ASUS or ABIT might be able to build new super servers with. And as another poster already pointed out, SUN already has large Opteron servers. Hopefully this will allow these types of servers to become cheaper and more common, encouraging more software to be written to support them.

      --
      And as you tread the halls of sanity, You feel so glad to be, Unable to go beyond. I have a message, From another time..
    3. Re:Do not count out Sun by rubycodez · · Score: 2, Interesting

      but that's a high-speed networked cluster, not a shared memory one? A 32 or 64 way shared memory opteron cluster is a threat to Sun's big-iron Ultrasparc boxes, would likely outperform them

    4. Re:Do not count out Sun by twiddlingbits · · Score: 1

      The Sun Opterons will run Linux, thats not proprietary. The hardwre implementations of such multi-processor are proprietary unless they are use PCI-X as the backplane to interconnect the processors, memory and peripherals.

    5. Re:Do not count out Sun by twiddlingbits · · Score: 2, Insightful

      UltraSPARC processors are due for an upgrade in early 2006. It's called and UltraSPARC IV+ and should be performance competitive with the Opterons. The problem is if folks already have a lot of SPARC based software X86 architectures are not a real option for them as the software is not portable. Some vendors of SPARC solutions may prefer to NOT support the X86 for various reasons. Sun will still have a market.

    6. Re:Do not count out Sun by Somegeek · · Score: 2, Informative
      Quoth twiddlingbits:
      The Sun Opterons will run Linux, thats not proprietary.
      TFA is about hardware, not software. What OS Sun runs on their servers is irrelevant to this discussion.

      Quoth twiddlingbits yet again:

      The hardwre implementations of such multi-processor are proprietary unless they are use PCI-X as the backplane to interconnect the processors, memory and peripherals.
      I don't think that you understood the word 'proprietary' in this context. Sun's technology, as well as (IBM's, SGI's, Cray's, etc.) is something that they each independently developed solely for their own company's use. They do not share or sell this technology to third parties to enable those third parties to build their own large multi-way servers using that technology. That is why their existing solutions are called 'proprietary'. It has nothing to do with if the bus that they use to connect the components is based upon some published open standard, whether that be PCIe, HyperTransport, or one that you think defines 'non-proprietary'.

      The two companies listed in TFA have developed solutions specifically to allow third party companies, whether they be Dell or ASUS or someone else, to build and sell multi-CPU solutions without developing the CPU interconnect technology themselves. That is why these new solutions are not 'proprietary'. They are not kept only for the private use of the company that developed them.

      I hope that makes this point clear.

      --
      And as you tread the halls of sanity, You feel so glad to be, Unable to go beyond. I have a message, From another time..
    7. Re:Do not count out Sun by SuperQ · · Score: 1

      Except the fact that most Opteron based Sun machines were designed by Newisys.

      Look at these pictures:
      Sun V40z
      Newisys 4300

      Newisys was one of the first places to design Opteron based systems.

    8. Re:Do not count out Sun by Anonymous Coward · · Score: 0

      "Give it another few months and I'm sure Sun will have some server with an obscene number of opterons in it, if thier current direction is any indication ...."

      Yeah, they can rebadge Cray Red Storm boxes, and continue to play follow-the-leader ;)

  4. Imagine a... nah, too easy. by meringuoid · · Score: 4, Interesting
    Cheap shots about Gentoo and Doom 3 aside, this is cool to see. I imagine it warms the heart of a lot of us old AMD fanboys. Plus, with a bit of luck the extra volume will bring down the prices of the Athlon 64s we stick in our gaming boxen. Right?... Right?

    ... k, maybe not. Can't afford one anyway :-(

    --
    Real Daleks don't climb stairs - they level the building.
    1. Re:Imagine a... nah, too easy. by davejenkins · · Score: 1

      he extra volume will bring down the prices of the Athlon 64s we stick in our gaming boxen. Right?... Right?

      Sure, as soon as you stop using that horrible term 'boxen'....

    2. Re:Imagine a... nah, too easy. by Mercano · · Score: 4, Funny

      I imagine it warms the heart of a lot of us old AMD fanboys.

      Plus anything else in the room it sits in. Great solution to the heating oil crunch! These things ain't P4s, but 32 CPUs is 32CPUs.
      --
      #include <signature.h>
    3. Re:Imagine a... nah, too easy. by wernercd · · Score: 1

      wouldn't the correct thing to say be: these things aint p4's AND 32 cpus are 32 cpus makes more sense that way IMO

    4. Re:Imagine a... nah, too easy. by Peter+La+Casse · · Score: 1

      No, I think "but" is the correct word to use. If they were P4s, the natural assumption would be that they create lots of heat, but despite the fact that they're not P4s, 32 of these processors will still create lots of heat. To rephrase, the fact that they're not P4s might lead someone to assume that they don't create lots of heat, but in fact they do, because there are 32 of them.

    5. Re:Imagine a... nah, too easy. by Anonymous Coward · · Score: 0

      You must not be a native English speaker.

    6. Re:Imagine a... nah, too easy. by Anonymous Coward · · Score: 0

      Cheap shots about Gentoo and Doom 3 aside, this is cool to see. I imagine it warms the heart of a lot of us old AMD fanboys. Plus, with a bit of luck the extra volume will bring down the prices of the Athlon 64s we stick in our gaming boxen. Right?... Right?

      Boxen? You are the ghey. Join the navy already.

    7. Re:Imagine a... nah, too easy. by wernercd · · Score: 1

      no. I am. 26 years old and all of them as an American :) But I don't 'nit pick' my posts and I was typing on my Treo650. But if it makes you feel better to complain about my grammar while hiding as an AC... more power to you

    8. Re:Imagine a... nah, too easy. by Anonymous Coward · · Score: 0

      It would even be more correct with punctuation and capital letters in the appropiate places.

    9. Re:Imagine a... nah, too easy. by SleepyHappyDoc · · Score: 1

      Plus anything else in the room it sits in. Great solution to the heating oil crunch!

      I assume you're making a funny, but I've actually noticed an improvement in my home heating since i expaned my network to four machines with five processors. In the temperate climate I live in, I don't need to use the electric heat at all in the winter, as I had to before when I only had one PC.

      Plus my PC has a lot more functionality than my baseboard heaters. ;)

      --
      Stasis is death. Embrace change.
  5. Supercomputing gets some more Opteron Action by skitheboat · · Score: 1

    Why roll your own when you can get the pricing of mass produced processors and focus your value add on better software, interconnect and optimization? Scale it up and down. Bring on a practical desktop cluster for video rendering ang Foling @ Home!

  6. Big Iron? Uhhh... by iBod · · Score: 0, Flamebait

    I think somebody doesn't know what 'Big Iron' is if they think AMD chips have anything to do with it.

    1. Re:Big Iron? Uhhh... by kevin_conaway · · Score: 5, Informative

      Ehh, maybe. Normally "Big Iron" is associated with IBM but according to Wikipedia, the submitter is correct in using the term.

    2. Re:Big Iron? Uhhh... by $RANDOMLUSER · · Score: 1

      Why not? "Big Iron" doesn't mean ONE of anything anymore.

      --
      No folly is more costly than the folly of intolerant idealism. - Winston Churchill
    3. Re:Big Iron? Uhhh... by BJZQ8 · · Score: 5, Insightful

      I think this represents a fundamental shift in what "big iron" of the future will be. Instead of a few ultra-reliable, ultra-expensive processors, we will use masses of somewhat-reliable, cheap processors. The 64-processor clusters are just the beginning. Sony/IBM's Cell is a step in that direction; lots of little processors, rather than one big one. Big Iron is just what you make of it, after all, and ultra-reliability in practice doesn't have to mean an archaic architecture in design.

    4. Re:Big Iron? Uhhh... by Mr.+Underbridge · · Score: 1
      I think somebody doesn't know what 'Big Iron' is if they think AMD chips have anything to do with it.

      I think somebody (ie, you) is confusing what big iron *is* for what it *was*. Things have changed in the days of the Beowulf cluster.

    5. Re:Big Iron? Uhhh... by Bert64 · · Score: 1

      Big Iron has already been available in 64 processor configurations for many years, the sun E10000 is quite old now, and i`m sure there were similar sized machines much earlier.

      --
      http://spamdecoy.net - free throwaway anonymous email - avoid spam!
    6. Re:Big Iron? Uhhh... by bloosqr · · Score: 4, Informative
      Whats your definition of Big Iron?

      There is a Cray XT3 that runs at 15 Teraflops at Sandia and made out of 2ghz opterons and is currently the 10th fastest computer in the world. There is a similar machine over at Oak Ridge National Labs that runs at 14 Teraflops and is the 11th fastest computer in the world.

      In fact, those lowly AMD kids seem to also have their chips on the fastest machine at the Pittsburgh supercomputing center (ranked 33rd fastest computer in the world) and the US Army Research Laboratory (ranked 39th fastest) . The latter was actually being built by IBM for ARL, you know those guys who coined the term "big iron".

    7. Re:Big Iron? Uhhh... by gormanly · · Score: 1

      No. They. Haven't.

      No-one seriously wants to run a bank on a Beowulf cluster.

    8. Re:Big Iron? Uhhh... by Sique · · Score: 1

      And what about the Foundry BigIron series?

      --
      .sig: Sique *sigh*
    9. Re:Big Iron? Uhhh... by neclimdul · · Score: 1

      I think you're misreading the title. He ment that Big-Iron(mainframe companies) are opening to AMD. This will allow AMD to scale up their processors using the Big-Iron tech. Sounds good to me.

    10. Re:Big Iron? Uhhh... by Mr.+Underbridge · · Score: 1
      No. They. Haven't. No-one seriously wants to run a bank on a Beowulf cluster.

      Nor do they want to spend 100x the cost of a Beowulf cluster on a single big iron machine that has less processing power. Which is why companies like those in the story are bridging the gap by speeding up the latency of multiprocessor computations. So you use commodity chips, get great performance, lower prices, and don't deal with clusters.

      Despite your assertions, the days of classic big iron are over. DEC got bought and rebought years ago. SGI and Cray are both penny stocks, that's how well their markets are doing. IBM's essentially out of that market completely.

      These days, "big iron" generally means a rack full of blades. The market for overpriced machines from the likes of Cray is dead. And putting periods at the end of every word in your sentence doesn't change that.

    11. Re:Big Iron? Uhhh... by slavemowgli · · Score: 1

      Huh? Since when does "big iron" have anything to do with the manufacturer of the CPUs used? Sorry, but your attempt to be cool by sneering at AMD's failing, kid - you're only making yourself look like a fool who doesn't know what he's talking about. :)

      --
      quidquid latine dictum sit altum videtur.
    12. Re:Big Iron? Uhhh... by Anonymous Coward · · Score: 0
      The latter was actually being built by IBM for ARL, you know those guys who coined the term "big iron".


      Everyone knows Johnny Cash coined that term
    13. Re:Big Iron? Uhhh... by IckySplat · · Score: 2, Interesting

      Hate to break it to you ...

      I've worked for Banks, Airlines too
      The all had BIG iron in the server rooms

      These people are not doing lots of hard calculations
      They are moving large amounts of data...
      Clock speed doesn't matter that much in these cases.
      I/O bandwidth is king for these applications

      The fact that these machines are ULTRA reliable
      Is a really big deal for these companies.
      You can't just reboot you're entire banking platform
      to add disk or fix broken hardware.

      Processing power just isn't the point in these cases

      --
      Help! help!, the termites are eating my DRAM!!!
    14. Re:Big Iron? Uhhh... by iBod · · Score: 1

      Listen up!

      a) was I talking to you? a: NO!

      b) did I say it was a manufacturer thing? a: NO!

      c) Kid? Don't call me kid you fucking loser!

    15. Re:Big Iron? Uhhh... by BigBuckHunter · · Score: 1

      Unfortunately, opterons are not yet reliable (refering to your "ultra-reliable" bit). I don't know if it's just us, but 5% of the opterons that we ordered give the following... (machine names have been disguised to protect the innocent, my job, and homosexual whales)

      [root@XXXXXXXXXXXXXX-oracle-1 root]# CPU 0: Machine Check Exception:
      0000000000000004
      Bank 4: f60d200159080

      Message from syslogd@XXXXXXXXXX-oracle-1 at Thu Oct 13 16:38:41 2005 ...
      XXXXXXXXXX-oracle-1 kernel: CPU 0: Machine Check Exception: 0000000000000004
      813 at 000000005dfee510
      ernel panic:
      Message from syslogd@XXXXXXXXXX-oracle-1 at Thu Oct 13 16:38:41 2005 ...
      XXXXXXXXXX-oracle-1 kernel: Bank 4: f60d200159080813 at 000000005dfee510
      CPU context corrupt

      This occurs with both the HP and ASA machines, with the exception that this can be generated on HPs by having mismatched "processor card" PCB versions.

      Thank you for your time,
      BBH

    16. Re:Big Iron? Uhhh... by FromageTheDog · · Score: 1

      Actually, this is an old idea. Anyone here remember Thinking Machines? All those pretty lights!

    17. Re:Big Iron? Uhhh... by Anonymous Coward · · Score: 0

      Hate to break it to you, but most of the big financial institutions have huge compute needs because they are not just processing accounts for customers. They are running huge market analysis codes to guide their own institutional investments. Go find someone in the R&D half of the IT department and ask about blades, clustering, network interconnects, the convergence of NAS and SAN, and how much of the instititution's future depends on this. If your institution of choice has no R&D half to their IT, they probably aren't really running contemporary big iron in the first place. :-)

    18. Re:Big Iron? Uhhh... by Anonymous Coward · · Score: 0
      c) Kid? Don't call me kid you fucking loser!

      Oh grow up and stop swearing like a 14 year old idiot.

    19. Re:Big Iron? Uhhh... by Anonymous Coward · · Score: 0

      You are talking about ARL being the ones that had ENIAC, right?

      http://ftp.arl.mil/~mike/comphist/eniac-story.html

  7. Clusters vs. single servers by Wesley+Felter · · Score: 4, Insightful

    SeaStar and InfiniPath (and don't forget the XD1) are great for building non-cache-coherent clusters, but those are mostly useful for running specially-written scientific applications.

    Horus is used for building Opteron ccNUMA machines with one OS instance that can run any Linux or Windows apps. It's a very different solution for a different market.

    1. Re:Clusters vs. single servers by dAzED1 · · Score: 1

      yeah, not really sure what the writer was thinking. One can cluster boards built with *any* processor out there. Can even mix and match.

      Hasn't Beowulf been out there long enough for people to understand this?

    2. Re:Clusters vs. single servers by joib · · Score: 1

      Well, unlike most cluster interconnects InfiniPath is special in that it connect directly to the HyperTransport interface on the Opterons, without going via PCI(-e|-X). So it won't work on e.g. a Xeon.

      But yeah, from a user perspective Infinipath is just like any other cluster interconnect. Perhaps a bit lower latency, but nothing dramatic (as in order(s) of magnitude lower latency).

    3. Re:Clusters vs. single servers by Tough+Love · · Score: 1

      Perhaps a bit lower latency, but nothing dramatic.

      What is not dramatic about 1.3 microseconds half round trip latency?

      --
      When all you have is a hammer, every problem starts to look like a thumb.
  8. What's the big deal? by Anonymous Coward · · Score: 0

    Or...you could go to eBay and get yourself a 64 processor Sun E10K for the price of a handfull of PCs. This technology's only been around for, what, 10 years now?

    Opterons == "non-proprietary" Ok, cool. So are Sparc.

    1. Re:What's the big deal? by Anonymous Coward · · Score: 0

      But some silly companies actually buy new equipment (you know, like not from eBay), and a new 64-core Opteron system will probably be faster and cheaper than a new 64-core SPARC machine. Heck, at this point a 64-core Opteron might beat a 128-core SPARC.

  9. Links by Red+Flayer · · Score: 3, Informative

    Just wanted to point out that the link to Newisys is just a blurb stating that AMD is releasing the Horus chip, and doesn't really have anything to do with Newisys, other than the fact that a couple of the people behind the AMD Horus release used to work there.

    Oh, and the Horus link is a PDF whitepaper... please warn when a link points to a PDF.

    --
    "Trolls they were, but filled with the evil will of their master: a fell race..." -- J.R.R. Tolkien on Olog-hai
    1. Re:Links by Wesley+Felter · · Score: 4, Informative

      No, Horus is developed by Newisys, but the people who initiated it have moved on from Newisys to AMD.

    2. Re:Links by Anonymous Coward · · Score: 0

      it's hard to miss that pesky .pdf text on the end of those links for sure!!!

    3. Re:Links by Red+Flayer · · Score: 1

      So why does the link say that the Horus is an AMD product? Is AMD just distributing a Newisys product?

      --
      "Trolls they were, but filled with the evil will of their master: a fell race..." -- J.R.R. Tolkien on Olog-hai
    4. Re:Links by m50d · · Score: 1
      Oh, and the Horus link is a PDF whitepaper... please warn when a link points to a PDF.

      Why? You can set up your browser to handle a pdf link any way you like (if it bothers you that much, you can redirect yourself to a google html conversion of it).

      --
      I am trolling
    5. Re:Links by Red+Flayer · · Score: 1

      Unfortunately, the PHBs here at the office disagree.

      --
      "Trolls they were, but filled with the evil will of their master: a fell race..." -- J.R.R. Tolkien on Olog-hai
    6. Re:Links by Chris+Burke · · Score: 1

      Because Le Inq got it wrong, or at least used confusing wording. Given the incestuous relationship between Newisys and AMD, this isn't surprising, and in fact referring to it as an "AMD product" isn't 100% off the mark... But the .pdf (and history) make it clear that it is a Newisys product.

      --

      The enemies of Democracy are
    7. Re:Links by macshit · · Score: 1

      Unfortunately, the PHBs here at the office disagree.

      Man, I've heard of bizarro office rules, but this takes the cake...

      Just out of curiousity, what happens when you download a PDF file? Do alarms go off and guards rush out to escort you away? [And what about .dvi files??? Robot machine-gun nests start pumpin' out the lead?]

      I'm just trying to imagine how your boss phrases a reprimand for something so innocuous... "Sorry Bob, but we're all insane idiots here -- you're fired."

      --
      We live, as we dream -- alone....
    8. Re:Links by Heretik · · Score: 1
      please warn when a link points to a PDF


      I've never understood how people are actually so retarded they can't look at the status bar to see where a link goes before clicking it. Because of your incompetence, the article should be littered with a bunch of redundant, annoying little warnings?



    9. Re:Links by Red+Flayer · · Score: 1

      Our browsers are reinstalled weekly (!).

      If the reinstalling tech notices anything out of the ordinary, a ticket goes in, and it gets brought up in weekly progress meetings.

      Not a tech company -- the PHBs are paranoid about 'hackers' because of an intern who left all sorts of malicious code in the system last year. He was using the office manager's account, which has root access... another problem. But since the office manager (the guy who orders the pencils) also supervises the IT team... well, it is what it is. And the default access for the office manager is root access...

      --
      "Trolls they were, but filled with the evil will of their master: a fell race..." -- J.R.R. Tolkien on Olog-hai
    10. Re:Links by Red+Flayer · · Score: 1

      I CAN look at the status bar... but that's inefficient. Looking all around my screen to find relevant information is disruptive, far easier to have "(pdf)" after a URL. And how badly does that really litter your screen?

      --
      "Trolls they were, but filled with the evil will of their master: a fell race..." -- J.R.R. Tolkien on Olog-hai
  10. god damnit by Douglas+Simmons · · Score: 5, Funny

    You kids, with your ultrasparc risc processing synchronous hypermultithreading vax/vms redbox pbx mumbo jumbo and your Ska music. For Christ's sake, cut the cotton-pickin' bullshit and tell me which stocks to buy and which to short. Oh and that AMD "capturing" the retail market tip the other day? Thanks for costing me six thousand dollars, my wallet was too thick and giving me a bad back. Christ.

    1. Re:god damnit by mikael · · Score: 2, Funny

      In my day, we used to toggle the OS and each program in bit by bit, remembering each opcode from memory. A keyboard was a musical instrument, a mouse was a type of vermin, and a terminal was where you attached the grounding wire of a lightning rod.

      --
      Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
    2. Re:god damnit by sconeu · · Score: 2, Funny

      You had lightning rods? We just told the tallest person... "Go stand in that there field!" when a storm came! And we LIKED it that way!

      --
      General Relativity: Space-time tells matter where to go; Matter tells space-time what shape to be.
    3. Re:god damnit by Anonymous Coward · · Score: 0

      What are you, like 4'6"?

    4. Re:god damnit by Anonymous Coward · · Score: 0

      Thats what happens when you buy High. Any chart choulda shown you that. Doesnt matter, if you were smart and held on to it, it will go up a whole lot more than that. Just give it time for Fab36 to pump out products. It just went online, so it generally will take a few months to make anything significant happen.

    5. Re:god damnit by Douglas+Simmons · · Score: 1
      Wake up and smell the Red Bull. It's time to get out of tech sector, son.

      And turn down that Ska music.

  11. The Topic is Misleading! by Anonymous Coward · · Score: 1, Funny

    For a minute I was excited I might get to pop an Opteron into my PDP.. that thing really lags when you try and load up Wolfenstein. Or "Adventure" for that matter.

  12. SMP memory model? by Anonymous Coward · · Score: 0

    Anyone know what the memory model for Horus is? Hopefully it isn't processor consistency type 2 which is all of the disadvantages of the TSO memory model with none of the advantages.

    1. Re:SMP memory model? by Wesley+Felter · · Score: 2, Informative

      It's the same as the regular Opteron (otherwise stuff would break).

    2. Re:SMP memory model? by Anonymous Coward · · Score: 0

      Actually, that's why I asked. Nobody is quite sure what the Intel and AMD memory models are. We weren't even able to get a direct answer from Any Glew on this. But we'll find out if we guessed right.

    3. Re:SMP memory model? by Wesley+Felter · · Score: 1

      Maybe you should ask Linus Torvalds and friends; they're the ones who have to worry about it.

  13. Is 32 by Andy_R · · Score: 1, Interesting

    really a 'large number'?

    --
    A pizza of radius z and thickness a has a volume of pi z z a
    1. Re:Is 32 by Anonymous Coward · · Score: 0

      I don't know... I've always fancied the following scheme for counting
      0,1,2,many, ...,many many. So decimal 32 would probably be in the many or many many bin and yes that would be a large number.

    2. Re:Is 32 by hcob$ · · Score: 1
      really a 'large number'?

      compared to 2.1230753x10^-32
      --
      Cliff Claven
      K.E.G. Party Chairman
      Founding Leader of: Koncerned for Egalitarin Governance
    3. Re:Is 32 by nganju · · Score: 5, Funny


      For this context, 32 is plenty large. Large is relative. If you ask me how many grains of rice I ate last night, 100 would be a small number. If you ask the average slashdotter how many women he's dated, 1 is a huge number.

      --
      There are 2 kinds of people in this world. Those that can keep their train of thought,
    4. Re:Is 32 by ArsonSmith · · Score: 2, Funny

      If you ask the average slashdotter how many women he's dated, 1 is a huge number.

      And the date was probably huge too.

      --
      Paying taxes to buy civilization is like paying a hooker to buy love.
    5. Re:Is 32 by IckySplat · · Score: 2, Funny

      Are those women real?
      Or integer /me ducks

      --
      Help! help!, the termites are eating my DRAM!!!
    6. Re:Is 32 by brunson · · Score: 2, Funny

      I have personally dated 14+32i women, and trust me... it can get pretty complex.

      heheheheheheheheheh... math humor... heheheh heheheh heheheh

      --
      09F911029D74E35BD84156C5635688C0
      Jesus loves you, I think you suck
    7. Re:Is 32 by Phishcast · · Score: 1
      Good work reinforcing the GP:

      "If you ask the average slashdotter how many women he's dated, 1 is a huge number."

    8. Re:Is 32 by GweeDo · · Score: 1

      It is so sad that this actually made me laugh...

      My poor wife.

    9. Re:Is 32 by aminorex · · Score: 1

      32 is a large number for SMP CPU heads, but it gets bigger. Consider that in roughly the
      same timeframe that retail chassis are available using these chipsets, quadcore CPUs
      will be on the market. That puts 128 modern CPUs in one box. Rare indeed is the whitebox
      application that needs to scale beyond that. And if you cluster *those*, you're rapidly
      approaching the top end of existing systems for HPC throughput.

      --
      -I like my women like I like my tea: green-
  14. Compiler technology by grahamsz · · Score: 1

    My experience suggests that Suns compilers beat out GCC on single processor machines and more so on 8 and 16 processor systems.

    The benifit of Sun's mature sparc compilers might let you squeeze more performance out of a sparc box. Although of course the opterons have a higher clock speed in the first place.

    1. Re:Compiler technology by Slashcrap · · Score: 2, Interesting

      My experience suggests that Suns compilers beat out GCC on single processor machines and more so on 8 and 16 processor systems.

      I didn't realise that the compiler was involved in optimising code for multiple processors. I thought it was the programmer that had to do that. I must have been misinformed.

      The benifit of Sun's mature sparc compilers might let you squeeze more performance out of a sparc box.

      So are you comparing the Sun compiler and GCC on Sparc or Opteron because you don't make that clear? One would expect the Sun compiler to do best on Sparc since nobody else has ever been that interested in it.

      If on the other hand, you are claiming that Sun have suddenly produced the best compiler for AMD's Opteron then I would like to see some more evidence.

    2. Re:Compiler technology by jedidiah · · Score: 1

      In the last shop of mine where 12+ cpu machines were used, gcc and friends were very much prefered over the Sun's payware toolchain. It wasn't cost either since even Sun's payware toolchain is dirt cheap compared to enough licenses to run Oracle on large NUMA machines.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    3. Re:Compiler technology by Anonymous Coward · · Score: 0

      A chip's clock speed means absolutely nothing, when comparing performance, unless you are talking about different revisions of the same chip, on identical motherboards.

      Example:
      A 2Ghz Opteron is faster than a 1.8Ghz Opteron

      But a 2Ghz AMD 3200+ is much faster than a 2.53Ghz Pentium IV, especially for IO bound process.

      Then there is the role that the system board architecture plays. If a FSB on a Pentium IV 2.53 Ghz is 800Mhz, and that's pitted against a 533Mhz FSB, *with the same CPU*, the 800 Mhz FSB system will be much faster.

      CPU Mhz is a very poor way to judge performance of a chip, much less an entire system, when comparing it against an architecture that is nothing (I mean NOTHING) like the one you are comparing it against. Apples and Oranges....

      About the only thing they have in common is word length.

      l8,
      AC

    4. Re:Compiler technology by Shinobi · · Score: 1

      Yeah, well, the whole GCC thing usually revolves around religious adherence to the only compiler most comp.sci weenies of today has ever used.

      I meet way too many comp sci people of today, quite a few of them extreme open source advocates, who can only work with GCC, and are very poor when it comes to writing ANSI-compliant code. They are too used to using GCC and the non-standard extensions. Quite ironic, given their pretentious drivel about "standards compliance"

  15. How about 2560 Opterons? by oringo · · Score: 1

    China has had that system running since June 2004: http://www.top500.org/sublist/System.php?id=7036
    Which makes my wonder, how much of the post is really news?

    1. Re:How about 2560 Opterons? by imsabbel · · Score: 1

      there is a HUGE different between a cluster and a shared memory system.
      HUGE.

      --
      HI O WISE PRINCE. WHT TOOK U SO DAM LONG?
    2. Re:How about 2560 Opterons? by networkBoy · · Score: 1

      While I am an Intel fanboy, and thus don't like AMD per se, this is in fact a huge deal. Shared memory systems are the holy grail of high performance computing, but traditionally have been so expensive that smaller groups (mom and pop shops, Jr. Colleges, many universities) simply can not afford them. This will open up a new era of HPC and I am glad for it. Now if they will support the P4m and/or Xenon I would be hopping with glee.
      -nB

      --
      whois gawk date unzip strip find touch finger mount join nice man top fsck grep eject more yes exit umount sleep dump
    3. Re:How about 2560 Opterons? by whorush · · Score: 1

      dude, not trying to start a brawl, i think its time to switch teams. amd scales way better and gives way better performance/$,watt. the p4/xeon suck, the itanium is a dud and the 25W turion beats the centrino. sorry :)

    4. Re:How about 2560 Opterons? by fimbulvetr · · Score: 1

      Just wondering, what is it you like about Intel? I mean, you said you're a fan boy, right? So what is it that makes you like intel? Is it because they make superior processors? Do you like intel because amd is the competition and you dislike AMD?

      There's probably something I'm not seeing, so I'm just wondering what it is.

    5. Re:How about 2560 Opterons? by networkBoy · · Score: 1

      I have used AMD in the past and every time I've had issues with the system. More than once (countless times actually since the 80286 days) I've had the CPU simply fail, and in the past I've seen support (hardware & driver) being sub standard when compared to the Intel solution.
      While I agree that the driver support seems to have gotten better over time, and in fact it looks as if AMD _may_ be finally coming into their own, at least as a gaming platform, I've been burnt one too many times by AMD to go back and try it again.
      Besides, the Intel solution is still vastly superior for my application (video processing), and I play all of two games, both on a console.

      -nB

      --
      whois gawk date unzip strip find touch finger mount join nice man top fsck grep eject more yes exit umount sleep dump
    6. Re:How about 2560 Opterons? by Vancorps · · Score: 3, Interesting
      Since the 286 days I've had both AMD and Intel and I gotta say, in those days there was very little difference between processors. Fast forward to today and I wonder if you've even look at AMDs offerings in the last 3 years. Video processing? I stream live dvd quality feeds using Opteron processors for a simple reason. Dual Xeon 2.8ghz server could make 2 high quality dvd streams with minimal frame drops. Powerful yes, now compare it to the dual Opteron 1.8ghz the lowest of the low end. I can do 4 streams with the same level of compression. In both scenarios I'm using the same Osprey 230 PCI-X cards.

      So I'm a big curious what is so vastly superior? Are you using Intel compiled codecs on AMD machines when you did your testing? Did you even do any testing? I'll admit I had some trouble getting things running smoothly with the Opteron box but the end results speak for themselves; especially when you move over to the 64bit world with 64bit capture drivers the Opteron blows away anything Intel has put out to date. Of course Intel 64bit support is slow as all hell right now so I'm sure that will change in the near future.

      While you may have been burnt by AMD I will stick with them for the time being until Intel shows some signs of turning around their product offerings. I'm still curious how a processor has gone bad though. In my experience once you get back the first 90 days its smooth sailing regardless of manufacturer. Only reason I can think a chip would die later in life would be from a PSU failure or some sort of disruption. I've seen that happen, never just seen a cpu die though. Always some other component causing it.

      Of course this is getting off track from the article. The Opteron is very well suited for these large machines so I'll be curious how they perform in real environments like Oracle and DB2 setups. Opterons bandwidth improve the more processors you throw at it so it'll be intriguing to see the results.

    7. Re:How about 2560 Opterons? by networkBoy · · Score: 1

      Back in the earlier days I noticed that my AMDs ran hotter than Intel equivalent procs. Now that the tables are reversed, and opteron seems to be proficient for this stuff I may look at AMD when the time for a new round of PCs comes up (I'll build one of each and see what happens). Most of my work is render based, not live streaming, so I never looked in that direction. In fact the long ass pipelines that make the Intel a poorer choice for games work marvelously for my application. I have to admit to not trying an opteron based system, as my target device was to run a single thread app, which all but takes over the system anyway.

      As to the software, it is a combination of C(++) compiled with Intel optimizing compilers and P4 assembly code for two of the tightest loops in the app. I do not desire to re-write them (as I did not do so in the first place).

      -nB

      --
      whois gawk date unzip strip find touch finger mount join nice man top fsck grep eject more yes exit umount sleep dump
    8. Re:How about 2560 Opterons? by Vancorps · · Score: 1
      Well to that I would say that it might be worth a few hours of investigating if the time savings = a cost savings that is. Otherwise there really isn't a point to switching. I know all of our new servers here are Opteron based. They are vastly more stable. Our DBMS both Oracle and MS SQL runs much more smoothly for some reason. Might have been the fact that I knew more when I setup these systems. The Intel compilers are the major stumbling block right now although there are many codecs that have been optimized for AMD and more graphics apps such and Maya, Max, and Lightwave have AMD binaries which dramatically speed things up. In the streaming world there is virtual dub, mplayer, cinerra, and a few others that I have tried out on different platforms.

      In either case, I'd tend to say its always good to test in multiple scenarios especially if you have luxury of being able to test implement before you have to purchase all the new equipment.

    9. Re:How about 2560 Opterons? by WuphonsReach · · Score: 1

      AMD really became suitable once NVIDIA rolled out their nForce2 chipsets (AthlonXP CPUs) and gave VIA something to cry about. Prior to that, it was definitely more hit/miss for finding a good AMD chipset. Ever since the AthlonXP days, I've been plenty happy with the AMD side of the equation (less expensive, good performance, good chipsets).

      The last Intel box I built was a BX chipset (440BX?). I have (2) AthlonXP systems still running (one is a VIA chipset the other is an nForce chipset). Plus 3 Opteron systems running a variety of chipsets.

      VIA has gotten a lot better at their chipsets too (I used the VIA KT800 in the one Opteron system).

      As for video processing, one of my systems is a dual-CPU Opteron 246. Works fine and is speedy.

      --
      Wolde you bothe eate your cake, and have your cake?
    10. Re:How about 2560 Opterons? by Shinobi · · Score: 1

      Investigate Pathscale's compiler suite. For x86-64 they are the fastest ones around that are still easy to use. For EMT64, the Intel compilers are still better.

      I don't regret the purchase, much as I don't regret the purchase of the Intel compilers or the IBM compilers for PPC either. All are infinitely preferable to GCC if you can make back what the compilers cost. They all have better ANSI compliance for example, and much better performance in the scientific apps I do.

    11. Re:How about 2560 Opterons? by hackstraw · · Score: 1

      So I'm a big curious what is so vastly superior?

      I'm curious too what is different in your setup to be able to stream 2x more data with comparable systems. Albeit the 1.8GHz Opteron has a little more horsepower than a 2.8GHz p4, but that should have little to nothing to do with video streaming. Is the Osprey in a pci-x slot on both machines? Are the memory subsystems similar?

      I'm just very curious. Because 2x performance on similar machines has to be explainable beyond the processors that are almost the same.

    12. Re:How about 2560 Opterons? by Vancorps · · Score: 2, Interesting
      Both machines the Osprey was used in a PCI-X slot. The memory ironically enough on the Dual Xeon was PC3200 ECC and PC2700 ECC on the Opteron box because the 1.8 is the only one that uses such slow memory. The difference being that the Opteron has much more bandwidth to utilize for ram but thats not what video processing tends to use. Block sizes greatly effect performance. I set the Opterons to write in 8meg chunks and it seems to save IO. The Xeons don't handle chunks that large for some reason. Perhaps because of how the processors share the bus so they inherently choke the machine under a full load.

      I was curious too, perhaps the AMD optimized apps were better optimized that the Intel ones? Beyond that its also the fact that these Xeons could only do 32bit while the Opterons were running happily in 64bit. The biggest pain there was finding capture card drivers that worked and were natively 64bit. Viewcast makes that difficult, thankfully the Linux drivers worked.

      I also might add that the test was a Dell PowerEdge 1750 vs an IBM eServer 325 and later the 326 with dual core which was even nicer. So Dell might just make slow servers while IBM makes em fast. I know I'm fairly impressed with the IBMs over the Dells. Many situations the Dells were choking and I replaced them with the IBMs and all was then smooth. Our webserver was a prime example. 72 million hits in a month the Xeon servers were maxing out. Replaced them with Opteron based IBMs and 186 million hits later they are barely breaking a sweat. 35% cpu with 32bit Windows Server 2003. I'm curious how the 64bit version will effect it since we're expecting 250-300 million hits in January. Going to be a good test.

      In any case, the Opterons have proved themselves around here. I had to fight for the first one and now all the owner of the company wants to hear is that there is an Opteron in the server he is buying. The systems are rock solid stable and even faster than I expected.

      Another thought I had about the reason the servers performance might be different. The Dells had 3 10k rpm ultra 320 scsi drives in a raid 5 configuration while the IBM just has two SATA 150 drives in raid 1. So perhaps the raid configuration was slowing down the write times.

      Granted, not an ideal benchmark but its still pretty impressive the amount of power I can shove into a 1U server using the Opteron.

    13. Re:How about 2560 Opterons? by hackstraw · · Score: 1

      Granted, not an ideal benchmark but its still pretty impressive the amount of power I can shove into a 1U server using the Opteron.

      This is better than a benchmark, this is real work.

      I asked for you to clear things up, and you bring up more variables into the mix.

      If you could, please clarify the following:

      1) you mentioned dual core opterons. Was your 2x performance noticed on both the single and dual core?

      Block sizes greatly effect performance. I set the Opterons to write in 8meg chunks and it seems to save IO. The Xeons don't handle chunks that large for some reason.

      2) How didn't the xeons handle this? Did the app crash? If so, this could be due to different default limits between a 32bit and 64bit system in terms of stacksize or something similar. Offhand, I don't see why an 8meg chunk would not work well on an xeon but work well on an opteron. (I obviously know nothing about your particular application, and I've never done such video streaming before either. Also, xeons nor opterons have that much onboard cache, so that could not seem to make a difference).

      The Dells had 3 10k rpm ultra 320 scsi drives in a raid 5 configuration while the IBM just has two SATA 150 drives in raid 1. So perhaps the raid configuration was slowing down the write times.

      3) That could make a difference, but my bet is with the scsi raid 5 and not the sata raid 1. I'm not sure why you need to write when your streaming info (cache maybe for later???). But if disk io is important for your application, I would expect CPU differences to be even less of an issue, but I'm still baffled. BTW, what is the cache difference on the drives? 8 megs is about a sweet spot in terms of cache on ide disks nowadays, did the scsi drives have less?

  16. So what? by goodmanj · · Score: 1

    PathScale has released InfiniPath which can be used with an Infiniband switch to make a high-bandwidth low-latency interconnect for a supercomputer cluster.

    This is news? We've been using an Infiniband-connected Opteron cluster for almost a year now. I got bids from half a dozen companies willing to sell us one. This is old, established tech.

    1. Re:So what? by joib · · Score: 2, Informative

      The special thing about InfiniPath is that the adapter is not a PCI-(e|X) card but rather connects directly to the HyperTransport interface on the cpu (requiring a special MB with a "HTX" connector), giving slighlty lower latency than a normal IB adapter.

    2. Re:So what? by moro_666 · · Score: 1

      This chip will make it possible to build 32 socket (64-core) shared memory Opteron systems.

      hypertransport or pci-e|X or whatever connection you may have my friend. 32 cpu's running at 1800mhz or more are so hard to synchronize that imho this whole "megayhpe" is much talk about much nothin. yeah 32 cpu's might be a good idea, but the only occasion you really want a smp machine is to use software that uses shared memory and shared i/o access ... but 32 cpu's begin to beat eachother down so hard that you barely have the power of 1cpu * 24, if even that. you spend alot time synchronizing and very much less than you'd expect "really working".

      this is the reason why "in which ever connected" beowulf machines are actually very good, they have usually special software (unless it runs in openmosix or openssi), that is exactly meant to avoid the bottleneck mentioned above. they all have independent memory access and they all have independent access to their own caches and swap. and this is very very very big time difference. if you run a "tiny" 4-way smp machine, then if one cpu is writing to memory that is shared between the 4 cpus then the 3 others cant read nor write to the same region (for obvious reasons). and for most applications this beats down the very basic idea of being run in a smp box anyway. cpu's sitting duck behind eachother. forking applications however dont benefit anything from sitting in one box and therefore it's better for them to sit in a separate machine at all, to avoid all kind of inter-machine latencys.

      this "great invention" is about as good as attaching 32 F1 formula cars behind eachother ... no they will not drive 32x(350km/h) (okay maybe its a bit too drastic but you get the idea).

      --

      I'd tell you the chances of this story being a dupe, but you wouldn't like it.
  17. Proprietary? by johansalk · · Score: 1

    Operton is still proprietary, isn't it?

  18. threat to big iron by rubycodez · · Score: 2, Interesting

    A 32-way SMP dual-core opteron box is a serious threat to Sun Enterprise boxes with 64 to 128 UltraSparc, even the hardware partitioning doesn't mean as much when you can just use two or more x86-64 boxes at probably less than half the price. For that matter, it also attacks HP's "superdome" Itanium2 servers and some of IBM's Power5 and Power6. The closed architectures and the proprietary Unix(tm) they run are in deep doo-doo

    1. Re:threat to big iron by Anonymous Coward · · Score: 0

      SPARCs are just about the least proprietary processors going; they're certainly less so than Opterons. Any company that wants can make SPARC processors, and historically the fastest SPARC processors aren't usually the ones being offered by Sun.

    2. Re:threat to big iron by rubycodez · · Score: 2, Interesting

      Eh, I have a huge library on the opteron timing, opcodes and instruction set, and machine language coding practices that AMD sent me free of charge. But there's plenty about the Sparc III that Sun is holding close to its chest, just ask any openbsd developer.

    3. Re:threat to big iron by brunson · · Score: 1

      With AMDs commitment to support the Xen hypermanager calls in their next chips, the partitioning advantage that Sun has will be drifting away, soon.

      --
      09F911029D74E35BD84156C5635688C0
      Jesus loves you, I think you suck
    4. Re:threat to big iron by flaming-opus · · Score: 1

      It takes a lot more than some core logic chips to build a big server, and takes more than a big server to acquire any market share. Look at the industry. Sun still dominates the Unix server market with expensive machines full of slow processors. Mainframes are still a billion dollar business.

      While I'm excited to see the possibility, you're not going to get anyone to spend this much money on a 64-way opteron box until they have been on the market for years, have been tested, and tried, and have lots of software vendors lining up behind them. Don't believe me, then ask why unisys doesn't have more market share. Substitute xeon for opteron, and that's basically what they've been trying to sell for the last six years. The only thing opteron has that the es7000 doesn't, is market hype.

      The real change is the reduced number of people who need 32-way boxes, or any sort. How many people are really outgrowing an 8-core opteron/xeon box? That number keeps shrinking.

    5. Re:threat to big iron by hyc · · Score: 1

      All these comparisons to Sun are pretty funny, considering that Sun sells Opteron machines, and their previous generation of Opteron systems were designed by Newisys. Those Newisys guys clearly have solid experience designing around the Opteron. I don't think it will take long before platforms built around Horus prove themselves worthwhile, and probably Sun and other AMD builders will adopt it en masse.

      Since Sun is already in the Opteron camp, none of this news hurts them. With the cheapness of x86 hardware, they can spin out Opteron workstations and servers at a fraction of the cost of their lumbering Sparc solutions, and the difference is pure profit. Even if it hurts Sparc sales, they'll be crying all the way to the bank.

      --
      -- *My* journal is more interesting than *yours*...
    6. Re:threat to big iron by segfaultcoredump · · Score: 1

      Uh, Sun's domain based partitioning and Xen/ESX have very little in common and are actually addressing two different problems. They are not replacements for each other, and can actually be complimentary if you combine them.

      In the world of server and process consolidation, there are 4 or so 'levels'. Sun's partitioning addresses 1 of these levels. Xen/ESX/UML addresses another. (Note that the cost per hosted OS/Application goes down as you go down the levels. )

      At the top level, you have things like Sun's hardware domains. They provide _dedicated_ (but dynamically reallocatable) hardware to a domain. The advantage is that a bad memory chip or cpu in one domain does not impact any other domains. The disadvantage is that it is an all or nothing type of allocation. You cant give 1/4 of a cpu to a domain (Sun's setup is even worse, you have to allocate a uniboard at a time, or 4 cpu's and its attached memory). There is no performance penalty for this setup.

      At the 2nd level, you have things like Xen, VMWare's ESX, UML, some Mainframe partitioning schems and such. Each domain/partition gets to run its own OS, total independant of other domains/partitions on the system, so instances are isolated from (most) software faults. However one bad CPU will still bring the entire system down. You do incur a performance penalty as you go through the virtualization layer (the amount depends on what you are doing. It can be as small as 1% or as large as 500%)

      At the third level you find things like BSD Jails, Solaris 'Zones' and the venerable 'chroot'. You have namespace isolation, and varying degrees of process isolation. But everything is still stuck on the same kernel. Performance is very good, but you loose flexability.

      The 4th level is the simple process of running to apps on the same box within full view of each other. (the unix world has been doing this for years, but this is a new thing for the Windows folks :-) )

    7. Re:threat to big iron by hackstraw · · Score: 2, Insightful

      A 32-way SMP dual-core opteron box is a serious threat to Sun Enterprise boxes with 64 to 128 UltraSparc, even the hardware partitioning doesn't mean as much when you can just use two or more x86-64 boxes at probably less than half the price.

      Would you buy that from Mom or Pop?

      A Sun Enterprise system is a system that is supported by the vendor for its OS (Solaris) and the hardware as well as other players, as the other systems you mention.

      A 32-way SMP dual-core opteron box is just a figment of your imagination without a corresponding operating system to match. (You would have gotten another +1 if you had mentioned Linux).

      My point is that there is more to a system than just the hardware. Sure, a 32-way SMP dual-core opteron box might be a great (overpriced and expensive) web server and a nice high end box for number crunching, but most of the larger systems are part of critical systems that cannot fail or garble your data. A 32-way SMP dual-core opteron box sounds great, but offers nothing to those whose reputations and careers depend on the system fulfilling its job.

    8. Re:threat to big iron by Jherek+Carnelian · · Score: 1

      also attacks HP's "superdome" Itanium2 servers and some of IBM's Power5 and Power6. The closed architectures and the proprietary Unix(tm) they run are in deep doo-doo

      Both the HP superdome and most IBM Power systems run Linux too.

      The value of the "big-iron" unix systems is not really in their CPUs, it is in their chipsets, which Horus does compete with, but also redundancy features like ECC and chipkill, redundant power systems, etc. Once you start adding all the equivalent featurues to a horus-based opteron system, you start to end up in the same pricing ballpark.

    9. Re:threat to big iron by Anonymous Coward · · Score: 0

      "With AMDs commitment to support the Xen hypermanager calls in their next chips, the partitioning advantage that Sun has will be drifting away, soon."

      Yeah, except Sun's virtualization isn't really that great compared to some of the competition.

  19. Strange use of "open" by xoboots · · Score: 2, Insightful

    Proprietary is proprietary. AMD chips are no more "open" than any other vendor's chips.

    1. Re:Strange use of "open" by Anonymous Coward · · Score: 0

      You don't understand.

      Proprietary meant, in this context, that the companies that develop interconnect chip don't sell them: they just sell complete systems. So if your motherboard manufacturer (ASUS, IWill, Tyan, ...) wants to enter the market (and compete with, say, SUN, SGI, IBM, HP, ...), they have to develop the interconnect chip.
      Now, with the availability of the HORUS, they can just buy one off-the-shelf.

      This makes a HUGE difference !!!

      Stephane

    2. Re:Strange use of "open" by xoboots · · Score: 1

      Actually, I quite well understand. You don't understand that you can't change the meaning of a word to suit your particular want-of spin-doctoring. "Commercial, off-the-shelf" != Open.

      The AMD chip is still proprietary as are the interconnect technologies. The fact that they come in packages you can assemble yourself to build your own custom solutions is really the focus of the article, but completely besides-the-point in terms of "openness". I don't mind the article or the technology but I don't see the need to frame it as "open".

    3. Re:Strange use of "open" by Vancorps · · Score: 1

      Sorry but what would you consider an open processor then? All the documentation on the Opteron is available free of charge. What more are you looking for them to provide?

    4. Re:Strange use of "open" by Anonymous Coward · · Score: 0

      Proprietary is proprietary. AMD chips are no more "open" than any other vendor's chips.

      Can you site where the articles or summary say that AMD chips are open?

    5. Re:Strange use of "open" by xoboots · · Score: 1
      Sorry but what would you consider an open processor then? All the documentation on the Opteron is available free of charge. What more are you looking for them to provide?

      Well, the Win32 API is documented, too. Is it open?

    6. Re:Strange use of "open" by Vancorps · · Score: 1

      You didn't answer the question. Everything you need to build an Opteron and associated compilers is available from AMD free of charge. So I don't see how it isn't open in any way. As I asked before, what more are you looking for them to provide you before you will call it "open?"

    7. Re:Strange use of "open" by xoboots · · Score: 1

      Nor did you respond to my comment which is obviously in a similar vein. That said, I wasn't aware that AMD supplied schematics and didn't protect their patents so that I could use their designs in my own products. The fact is, I can not go out and implement my own Opteron's nor use their designs without licensing.

      I don't like to confuse the fact that the API/ABI is public with the notion that the product itself and all its workings are non-proprietary -- because they are proprietary.

      None-the-less, I'll concede that they divulge a lot of information but that does that not mean that they are not proprietary (which the OP was contrasting). I hate to do this, but here is a snippet from the wikipedia entry for "proprietary":

      "Something proprietary is something exclusively owned by someone, often with connotations that it is exclusive and cannot be used by other parties without negotiations. It may specifically mean that something is covered by one or more patents, as in proprietary technology. It can also mean that the copyright is used in a way that restricts the users' freedoms. "

    8. Re:Strange use of "open" by Vancorps · · Score: 1
      Well the Opteron isn't exclusively owned by AMD. It is based on technology from other companies with a few patents that AMD developed. It sounds to me like the word open is being confused with free. I see an inherent difference. AMD puts a lot of R&D into the product which would not exist without recouping costs and funding further R&D. So yes, the schematics are available and in many cases readily online so you could reproduce the Opteron without asking AMD as long as you didn't intent to profit from it.

      I don't see anything wrong with this and I would call it open.

      As for your comment I didn't feel I needed to respond to it as the discussion wasn't about Microsoft and its proprietary nature. There is no debate there is a lot of the platform that is either not documented or not publically available. Thus I would agree it conforms to the definition of proprietary.
    9. Re:Strange use of "open" by Anonymous Coward · · Score: 0

      From that definition, the Linux kernal is "proprietory", since it is copyrighted and licenced by the GPL.

    10. Re:Strange use of "open" by xoboots · · Score: 1

      I didn't say there was anything "wrong" with it -- I said it wasn't "open" and that using that term is both unneccessary and incorrect. If you can't reproduce it freely or build on it freely then it is not open. This is not strictly about freedom either, but about the meaning of "open" in the IT world. If you think "open" means you get to read their books then I guess your definition would work. I say it means I get to *use* what I read how I want to use it. It is not merely being able to build my own processor for my own use in my own living room.

      Anyways, I guess we'll have to agree to disagree. Best regards to you. ...but to the AC who says that linux is proprietary by this definition of "open" -- please go to the corner and beat yourself with a large clue stick.

    11. Re:Strange use of "open" by aminorex · · Score: 1

      That API is open. It's implementation is closed, and many auxilliary APIs are closed.
      Open means you can see inside. I can see inside the API, but not inside it's implementation.
      By publishing books on the subject, Microsoft has made the design open, to the degree
      that it is accurately documented, but not the code. OpenSolaris is open at the code level,
      but not at the process level: It does not use an open development process.

      I think this is a common jargon use of the term "open", and it corresponds closely to its
      dictionary definition.

      --
      -I like my women like I like my tea: green-
    12. Re:Strange use of "open" by csirac · · Score: 1

      Sorry but what would you consider an open processor then? All the documentation on the Opteron is available free of charge. What more are you looking for them to provide?

      I'm not sure that Opteron is necessarily more "open" than Sparc. NEC and Fujitsu (and probably others) sell sparc-based machines.

      The fact that it hasn't attracted the attention of consumer-level manufacturers has nothing to do with its "open-ness"...

      If there's something specific about Opteron that is more "open" than sparc, I'd like to hear about it.

      NB: Don't get me wrong. I wouldn't pay for a SMP Sparc box over an SMP Opteron one. Especially with their dual-core CPUs in each socket.

  20. On his hip? by Anonymous Coward · · Score: 0

    The swiftness of the processor is still talked about today...

  21. about time and huh? by Anonymous Coward · · Score: 0

    Well, it's about time people took the opteron seriously. But, it's not really designed for this. For best results it's a cpu designed for 4way (8 core) systems. They will need to do some finagling and commit some sins to get it to work, which almost always leads to an architecture that's less than optimal. This will lead people to believe that AMD isn't up to the job, since "oh yea, those AMD mainframes are garbage" when they don't get it to perform as well as the people building these things obviously think they should. I'd lay money that Intel is designing a 15 pipe interconnect for their 64 bit chips so they can swoop in and save the day when this idea fails.

    I think it might be better for AMD to come up with a new chip that has more interconnects, and design the chip for the job, instead of designing the job for the chip.

    Things almost always don't work right when you try to make them do stuff they aren't designed for.

    It's almost as if Intel put these companies up to this to generate FUD about Opterons... Maybe the Microsoft/Linux fiasco with the FUD has me suspicious of something innocent.

    There is no better marketing tool than to put the competition to a task it can't possibly do well and point saying "see I told you AMD is no good!".

    If I were Hector Ruiz, I'd publicly say that "..these people are doing this on their own and the opteron is not designed to do this, so it will never work well."

    l8,
    AC

    1. Re:about time and huh? by Cruithne · · Score: 2, Informative

      But, it's not really designed for this.

      And what makes you qualified to state this? Opterons were designed with EXACTLY this in mind, right off the drawing board - i'd dig up some old articles about it but I'm at work. Research Non-Uniform Memory Access (NUMA), supercomputing, and the Opteron's architecture as it related to those two. AMD knew what they were doing when they designed the Opteron - Intel has been completely out-engineered.

  22. Open Processors by Feneric · · Score: 2, Informative

    Umm, I know there's this odd phenomenon where many people tend to label any processor that's made by either Intel or AMD "non-proprietary" and any processor made by another company "proprietary", but even still this article is a little silly. SPARC processors have been in use since the late '80s, most people consider SPARC-based machines "Big Iron", and the SPARC processor architecture is fully open -- anyone who wants to can make SPARC processors. SPARCproductDIRectory lists a bunch of companies who currently do. In fact, there are probably just as many (if not more) SPARC manufacturers as there are X86 manufacturers.

  23. Word from the Mac community... by jnadke · · Score: 2, Funny

    This just in: Mac users say 64 core Opteron server will be almost as fast as the new Mac G5.

    1. Re:Word from the Mac community... by Anonymous Coward · · Score: 0

      MickeyMouseKetosh - Bletch!

    2. Re:Word from the Mac community... by manno · · Score: 1

      now that was funny, and honestly it's only about half as fast a the dual processor 2.7Ghz G5

  24. multithreading not pervasive... yet by GunFodder · · Score: 1

    AMD and Intel have introduced multicore chips to the consumer market. It won't be too long now before there are millions of home computers with SMP class processing power. I suspect we'll start seeing games that take advantage of multiple processors by the end of next year. Whether they scale past 2 cores is an interesting question.

    1. Re:multithreading not pervasive... yet by Hal_Porter · · Score: 1

      Well John Carmack has said

      http://techreport.com/etc/2005q3/carmack-quakecon/ index.x?pg=1

      "Graphics accelerators are a great example of parallelism working well, he noted, but game code is not similarly parallelizable. Carmack cited his Quake III Arena engine, whose renderer was multithreaded and achieved up to 40% performance increases on multiprocessor systems, as a good example of where games would have to go. (Q3A's SMP mode was notoriously crash-prone and fragile, working only with certain graphics driver revisions and the like.) Initial returns on multithreading, he projected, will be disappointing."

      Mind you if you could really get a 40% speed up it would mean that a X2 4800 at 2.4Ghz would outperform a FX-57 at 2.8Ghz (the fastest available dual core when the X2 4800 was released). In fact it should outperform anything up to 3.3Ghz.

      Actually half that, +20%, should be enough to level the playing field between the fastest available dual core and the fastest single core, assuming the 20% clock diff bewteen the X2 4800 and the FX-57 is typical.

      --
      echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
  25. From +5 Insightful to 0 Flamebait in 1 hour by iBod · · Score: 1

    Go slashdot!!!!!!!

    1. Re:From +5 Insightful to 0 Flamebait in 1 hour by Anonymous Coward · · Score: 0

      I can see it now ...

      meta-mod wars!

    2. Re:From +5 Insightful to 0 Flamebait in 1 hour by Anonymous Coward · · Score: 0
      c) Kid? Don't call me kid you fucking loser!

      It is because of this juvenile post by you.

  26. DEC, SGI, and Cray never made Big Iron. by Richard+Steiner · · Score: 1

    More to the point, supercomputers are not mainframes, which were (and still are, at least in most circles) the only systems that were traditionally associated with that term, and most of the folks who assert that mainframes are dead are not working for (and have never worked for) the companies who have traditionally used Big Iron in the first place.

    Those of us who still work in the airline, banking, or insurance industries know that most of the current mainframe systems aren't going anywhere. They run on Big Iron for a reason.

    Eventually they will be replaced, as everything is, but IMO their eventual replacements will be mainframes in everything but name -- ultra-secure, reliable, recoverable systems acting as large-scale corporate data servers.

    --
    Mainframe/UNIX Bit Twiddler and long time Windows/Linux Hobbyist.
    The Theorem Theorem: If If, Then Then.
    1. Re:DEC, SGI, and Cray never made Big Iron. by Vancorps · · Score: 1
      I happen to have a few friends in the banking industry and I find your statements curious. Are 100 clustered servers not more reliable than 1 mainframe? 100 clustered servers not more powerful than 1 mainframe? The price still isn't even close. Also, the banking industry has been clustering for quite some time now. It makes good sense considering database access works better inherently when you add more processors and more threads to the mix.

      After my latest experience with EMC equipment I have concluded that for the vast majority of applications clustering works better at a much better price point. So as your statement said, the replacements will be mainframes in everything but name. Ultra secure, reliable, recoverable systems. Around here we call it Oracle and the grid. Its real hard to take down the RAC here and I would say it is all safe and secure with the additional security found in the ability to have half the cluster here at HQ and the other half off site.

    2. Re:DEC, SGI, and Cray never made Big Iron. by Mr.+Underbridge · · Score: 1
      Those of us who still work in the airline, banking, or insurance industries know that most of the current mainframe systems aren't going anywhere. They run on Big Iron for a reason.

      But, since the whole point of this thread is sales, how many of those systems are being replaced? And how quickly? And is that enough to prevent the general demise of so-called "Big Iron" from a sales perspective?

      My guesses are not many, slowly, and no, respectively. There will always be niches for everything (even COBOL, God love it), but that doesn't mean it's not nearly dead for non-legacy applications.

  27. How is Opteron not proprietary? by Markus_UW · · Score: 1

    How is Opteron not proprietary? Only Intel and AMD can manufacture AMD64/EM64T chips. These chips (as with their x86 brethren) are among the most proprietary of chips, up there with Power 5. The only chip that's anywhere close to not being proprietary is the SPARC, which just about anyone can make (though only Sun and Fujitsu seem to).

    1. Re:How is Opteron not proprietary? by Anonymous Coward · · Score: 0

      How is Opteron not proprietary?

      Who said Opteron was not proprietary? Read the summary again. The paradigm is proprietary.

    2. Re:How is Opteron not proprietary? by default+luser · · Score: 1

      AMD64 / EMT64 / x86-64 is NOT PROPRIETARY.

      It was developed as an open standard to ensure industry-wide acceptance. Intel does not need a license to develop and sell an x86-64-compatible chip, and neither do you.

      According to Marc Miller of AMD: x86-64, now named amd64, was developed as an open standard, therefore it's not something we can license.

      The open standard is x86-64. AMD's trademarked implementation is called AMD64. Intel's trademarked implementation is called EMT64.

      --

      Man is the animal that laughs.
      And occasionally whores for Karma.

  28. Great AMD is quit is doing fine. by JollyFinn · · Score: 4, Insightful

    So where I can buy the AMD server with near full redundancy?
    Or the server which can run highly debugged application written in mainframe assembler in 60's or 70's ?
    Or atleast AMD computer with SINGLE memoryspace atleast 1TB in size?
    And also how many decades of uptime is for the operating system which is used with the new AMD computer?
    The horus is more or less getting close to midrange server in number of processor while it won't bring it to the reliability requirements of midrange server, to get that it would have to run its own memory controllers instead of cheap ass opteron controllers which lack for example hotswappable memory.
    Sure you get speed, but after taking the speed there is eventually a crash.
    The big iron is all about gettin continuing to function no matter what comes.
    Only problems outside of box, like earthquake or something similar could bring it down.

    Yeah. AMD is doing just fine...
    Its eating the cheap ass market, not the big iron.
    The price is cheap and its bought where the crash proof means better than windows which is like saying saying its unsinkable since it does better in open seas than normal rowboat used in lakes.

    Lets put it this way. x86 is just used in low end boxes and in clusters of lowend boxes. And those things are not for everything. They can do much but not everything. They are cost effective when you compare only the purchase price. But not so cost effective when downtime costs a lot.

    There is probably order of magnitude or TWO orders of magnitude of what joe slashdotter thinks big iron and what businesses have in big iron as in price range.

    --
    Emacs is good operating system, but it has one flaw: Its text editor could be better.
    1. Re:Great AMD is quit is doing fine. by JollyFinn · · Score: 1

      I wanted to add one thing.
      I'm having cheap ass Athlon 64 computer under my desk at home.
      And I think its great, but its still a very low end computer.
      And I think linux is a lot better than windows, and use it but its still no where near the mainframe OSes.
      Yes the mainframes have linux as one of their guest os. Running on top of a real os, but thats another story.

      --
      Emacs is good operating system, but it has one flaw: Its text editor could be better.
    2. Re:Great AMD is quit is doing fine. by manno · · Score: 1

      good point

    3. Re:Great AMD is quit is doing fine. by bunyip · · Score: 1

      So where I can buy the AMD server with near full redundancy?

      You don't, you buy two.

      Or atleast AMD computer with SINGLE memoryspace atleast 1TB in size?

      Two thoughts on this. 1. It's really, really expensive. 2. The mainframe assembler from the 60's or 70's is 24-bit or 31-bit addressing at most, 1TB or RAM won't help you.

      Only problems outside of box, like earthquake or something similar could bring it down.

      Now you put the second one 1000km away.

      BTW - I've replaced IBM mainframes with clusters of 4-way Opteron servers, so I know this works. And, over time, it will work more and more.

      Alan.

    4. Re:Great AMD is quit is doing fine. by bloosqr · · Score: 1

      The difference is no longer an order of magnitude or two even for the SMP boxes that you are talking about. Googling a bit, I have found these reasonably priced Opteron boxes. These are variations of the same opteron v40z boxes that Sun sells. Quad Opteron boxes w/ hot swap power supplies and hot swap hard drives w/ ECC memory for reasonably cheap for the chassis.

      I've never seen a terabyte of ram on a single domain but the
      very high end sun boxes will allow you to use either 576 or 288 gigs of ram on a single box, so I will use that as the base. The maximum in SMP mode for the opteron boxes you can homebrew is probably one based on the iwill motherboard which supports 64 gigs, which is now grabbing the midrange line of the sun server series.

      In fact HP makes the ProLiant DL585 with opteron chips. They will sell you this machine w/ up to 128 gigabytes of ram, redundant roms, redundant power supplies, hot swappable hard drives, ECC memory.

      Its not quite at the high end server sun range but the opteron chip and its use in machines built by traditional "big iron" builders is not really an order or magnitude off. Its really about a factor of 5 from the really high end (aka $1m+ sunfire installations).

      And as the IWILL MB points out its not even that far away from home brew or at the very least a $100k officebrew given the amount of ram one has to buy.

      Also its worth pointing out the processor supports up to a terabyte, the issue is the memory sticks/mb's are not available yet.

      And here we are only talking about the SMP/redundancy style computing . For distributed style computing platforms (aka everything on the top 500 super computers), the 10th and 11th fastest computers in the world are Cray machines based on opterons currently running at Sandia National Labs and Oakridge National Labs respectively.

      Now perhaps in your world the top 500 super computers are not considered big iron because they are distributed but i'm w/ the joe slashdotter crowd that considers "big iron" to include the top 500 fastest computers in the world.

    5. Re:Great AMD is quit is doing fine. by LWATCDR · · Score: 1

      "Or the server which can run highly debugged application written in mainframe assembler in 60's or 70's ?"
      Right here. Just load it up and go. Probably would run faster than it did in the 70s as well. http://www.conmicro.cx/hercules/

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
  29. Indeed. by turgid · · Score: 1

    If I were a betting man, I'd give you my trousers.

  30. Re:Not much to say about AMD, by Dahan · · Score: 1

    Hey, you lost your kanji .signature?

  31. NUMA, NUMA, yay! by tepples · · Score: 1

    Research Non-Uniform Memory Access (NUMA)

    Ma-ya-hee, ma-ya-hoo...

  32. Re:Not much to say about AMD, by Anonymous Coward · · Score: 0

    yeah, one day I logged on and my signature had been reduced to a BR tag. it makes me sad, but it also frees me up to change my user info -- in order to preserve the Kanji sig, I had to leave my old user info untouched.

    have fun at +2 you whore,

    --sa

  33. Some software just runs faster on Opterons by Anonymous Coward · · Score: 1, Interesting

    No why -- just how the instruction mix turns out. One example is PostgreSQL. Where Opterons might be 15% in MySQL, MS-SQL, DB2 versus a Xeon at the same price range, it's 100%+ faster than a Xeon in PostgreSQL. One of the things noted was large context switching was a big penalty on Xeons so there's been an effort to reduce it.

  34. Great interview with Horus' Engineers by algae · · Score: 1

    Real World Tech has a great interview between David Kanter and two of the engineers working on Horus. If you're interested in actual information about how it works and what it does, its good reading.

    BTW, Linus Torvalds posts to RWT :-D

    --
    Causation can cause correlation
  35. Obligatory quote by smartdreamer · · Score: 1
    Imagine a cluster of those!

    sorry...

  36. sweet! by Anonymous Coward · · Score: 0

    Oh fuck yeah! A paradigm change! I fucking love paradigm changes, and shifts too! This is fucking awesome, we should have a paradigm shift party! Fucking roxxor!