Cray Supercomputers to be Based on AMD Opterons

PsychicX writes "AMD and Cray have announced an agreement to base Cray supercomputers on AMD's Opteron line until the end of the decade, and to collaborate on Cray's 2006 proposal for Phase 3 of the federal government's DARPA HPCS (High Productivity Computing Systems) program. Cray already offers the XT3 and XD1 supercomputers based on Opteron."

Comments

excellent

[harryoyster]

That is excellent news for AMD even though there wont be massive volumes compared to home markets it will still be some heavy industry weight backing the AMD opteron processor. Hopefully AMD will adopt some additional features that could make the Opteron even better suited to the super computer market.

Re:excellent

[14erCleaner]

Except that this "industry heavyweight" is actually Tera, the little company that bought out the Cray name as the supercomputer industry was dying. Only the name Cray remains, not the old-time reputation.

Re:excellent

[Rude Turnip]

This is really just a marketing play on AMD's part. Now they can sell you:

"The SAME CPU used in CRAY ***SUPERCOMPUTERS***, now available for your desktop!"

And some rube will buy one based on that statement.

Re:excellent

[drinkypoo]

You mean kind of how they were trying to sell alpha workstations based on the Cray buzz when the T3 came out?

Re:excellent

[Anonymous Coward]

This is really just a marketing play on AMD's part. Now they can sell you:

"The SAME CPU used in CRAY ***SUPERCOMPUTERS***, now available for your desktop!"

And some rube will buy one based on that statement.

A well-scrubbed, hustling rube with some taste?

Re:excellent

[fgodfrey]

Tera bought far more than a name when they bought us. They also bought a bunch of software and hardware people, many of whom (myself not included) have been with Cray Research (the original Cray) for many years. So, while it's certainly not the Cray of the mid-1980's, the tradition still goes back there, especially with the vector machines like the Cray X1/X1E and its impending follow-on.

Re:excellent

[darkmeridian]

AMD has no incentive to change its server chip just to better suit the niche supercomputer market. It would only alter the Opteron to fit the server market, not the supercomputer market. The firm already has manufacturing bottlenecks; its production fabs are cramped as is. There are also problems with inventory. Why would AMD spend the money to create a chip that would benefit supercomputers and not servers? It would make sense only if AMD and Cray improved the tech in a way that would make the Opteron better suited to the server market as well, such as by improving the scalability of the technology.

Re:excellent

[Aardpig]

Only the name Cray remains, not the old-time reputation.

That's not quite true; they still sell Cray-specific technology. One of my colleagues has just bought a small 24-core Opteron system. Each node contains two dual-core processors, and the 6 nodes are linked together by 80 Gb/s Craylink cables. I think this interlink technology is also licensed to SGI for use in their Origin computers.

Re:excellent

[cameldrv]

They will make the appropriate changes. Look at the instruction set for any peculiar additions in the next rev of the opteron.

Re:excellent

[joib]

Tera bought far more than a name when they bought us. They also bought a bunch of software and hardware people, many of whom (myself not included) have been with Cray Research (the original Cray) for many years. So, while it's certainly not the Cray of the mid-1980's, the tradition still goes back there, especially with the vector machines like the Cray X1/X1E and its impending follow-on.


You work for Cray. Cool.

Please tell me that this deal implies that AMD is going to add some proper vector instructions to amd64. Pretty pretty please. ;-)

Specifically, I'm thinking about something like Alpha Tarantula. Well, the huge bunch of execution units and memory BW of tarantual was perhaps a bit overkill for a general purpose processor, but the vector ISA extensions and the general architecture looked all right. On paper, of course.

Re:excellent

[evilviper]

The firm already has manufacturing bottlenecks; its production fabs are cramped as is. There are also problems with inventory.

What bullshit. Provide some sources for this info, or shut up. AMD is opening a new fab, and has a contract with a 3rd party to produce cores if AMD can't keep up... They're doing fine producing Opeterons.

Incidentally, I can provide real, actual sources that show Intel is the one who is having problems producing enough chips to meet demand.

Re:excellent

[GuidoW]

Saying "I can provide real and reliable sources" without actually doing so is just as dodgy, maybe even more so, than not talking about reliable sources at all.

Re:excellent

[flaming-opus]

Which is true only insomuch as the old-time reputation could not possibly exist today. That was the cold war, this is the post-coldwar era. The old cray was a mammoth beast with its own share of myopia, but a lot of technical tallent. This allowed a few really brilliant concepts, and a lot of clever implementation to power two decades of brilliant computers. That said, they were brilliant solutions for the era. Old-school cray systems were built in an era when doing fundamental pieces of math was still pretty difficult, and the government was willing to pay ten million dollars for a machine that was proficient at doing math, and many tens of millions for a machine that was really good at it.

The difficult problems in building computers has changed, and the financial climate around supercomputers has changed quite a lot. Among other things, CMOS finally became fast enough to put bipolar in its grave, single microprocessor workstations became powerful enough to do all but the hardest of scientific tasks, and the average price of high performance (not top 10 on the list, but still fast) computers has plummeted. To ask the new cray to be like the old cray would be foolish.

That said, New Cray is still offers impressive products. All of Cray's 3 product lines have much lower entry-prices than similar crays of the 90's. They all have more managable power/cooling/physical size characteristics. They make much greater use of industry standard Disks and networks, and also can be administered and programmed much more like any other unix computer. You program a New Cray more or less the same as other contemporary HPC systems.

When cdc introduced the 6600, the president of IBM complained to his staff aking (paraphrase) 'how has cdc managed to best IBM's fastest computer with a staff of just 14 engineers and 4 programmers?' Seymor Cray responded "It seems like Mr. Watson has answered his own question." Because new Cray is tiny does not mean that it is not capable of making impressive innovations. Old Cray's Gorilla days were very wasteful, and not necessarily full of the best moments of innovation.

Now, if only they could put four X1e CPUs into an air-cooled, rack-mount server and charge a reasonable amount for it. I'd much rather have a handful of vector processors than a few dozen opterons, anyday.

Re:excellent

[HotNeedleOfInquiry]

A well-scrubbed, hustling rube with some taste?

A well-scrubbed, hustling rube with some taste and DOE/DOD dollars to spend.

Re:excellent

[joib]

Now, if only they could put four X1e CPUs into an air-cooled, rack-mount server and charge a reasonable amount for it. I'd much rather have a handful of vector processors than a few dozen opterons, anyday.


NEC sells an entry-level deskside SX-8, called IIRC SX-8i, with one processor. Unfortunately "entry-level" in this case means $100000+. :(

Re:excellent

[robertjw]

It would make sense only if AMD and Cray improved the tech in a way that would make the Opteron better suited to the server market as well, such as by improving the scalability of the technology.

Correct, but it would be perfectly reasonable to project that a company that is pushing the processor to the limits of it's performance, like somebody manufacturing supercomputers, would find areas for overall improvement of the chip.

Re:excellent

[convolvatron]

forrest, tera is finally fully dead, which you know full well.

so its all up to you now to make it happen, you have the
company to yourself. the success or failure of which will
be entirely of your own making.

Re:excellent

[masklinn]

Say, does on often open new factories when one's production is above one's sales?

The very fact that they need to open new factories and strike deels with other manufacturers to help them if they can't manage enough production DOES mean that their fabs are cramped and working full steam.

Re:excellent

[I Like Pudding]

One of my colleagues has just bought a small 24-core Opteron system

Small? Small? You've lost your sense of scale with all the weather modeling and nuclear RSA NP-hard protein folding simulations you've been running.

Re:excellent

[tzot]

...by 80 Gb/s Craylink cables. I think this interlink technology is also licensed to SGI for use in their Origin computers.

Which they call NUMALink (for Non-Uniform Memory Access --a node obviously accesses its memory faster than through the interconnection), and they're used in their Altix (Linux on Itanium2) computers too; I don't know how much they have evolved the technology, I just know that it's 3.2GiB/s each direction.

NUMALink

Have you got any links to some page describing some configuration similar perhaps to your colleague's system? I haven't heard about 10GiB/s CrayLink and find it very intriguing.

Re:excellent

[flaming-opus]

Yes True.
And NEC has done a great job of doing this for the last several generations of their vector machines. I have not ever programmed for an SX, and don't know much about them. The really nice thing about the X1, is that under the covers it's running Irix, which is a pretty reasonable Unix variant. Anyone know anything about super/UX?

Re:excellent

[tzot]

...and more research hints that "CrayLink" has little to do with Cray, apart for the name:

SGI at Wikipedia

So CrayLink and NUMALink are SGI developed technology, based on the DASH project at Stanford University.

Re:excellent

[flaming-opus]

Close.
Craylink was designed at SGI, and renamed to craylink after they bought Cray. They introduced craylink in the origin2000, which they started selling half a year after buying cray, so I'm sure they couldn't have integrated any cray-designs into their product in that span.

After they sold Cray to Tera, SGI started calling the technology Numalink, and currently use it in their origin3, altix3, and altix4 product lines. They are on the 4th generation of the technology, which is 3.2GB/s per direction. The cray that was sold to Tera included the half-finished X1 system, which also uses numalink. It uses the older 1.6GBps/dir links, but uses 32 networks in parallel for a total of ~50GB/s/dir per node.

The Cray XT3 uses a newer network interconnect called seastar, which offers 3.8GBps/direction. This is probably what will be used in the X1's successor.

The Cray XD1, which your colleague bought, is a product cray acquired when they bought OctigaBay. They use an interconnect called the RappidArray switch, which provides 4GBps/direction of interconnect.

All of these interconnects are high-bandwidth and low latency. The XD1, is also very inexpensive for a cray, which is always nice.

Re:excellent

[evilviper]

The very fact that they need to open new factories and strike deels with other manufacturers to help them if they can't manage enough production DOES mean that their fabs are cramped and working full steam.

No, it usually means they are planning ahead. Intel, in fact, is also opening new fabs, and that news, oddly enough, gets massive coverage on /.

Re:excellent

[flaming-opus]

No they won't! They have no reason to. The vector units that a cray uses aren't like altivec, sse, or other "bolt-on" vector units. The vector unit on a cray (or NEC) is a latency hiding mechanism. It's a method for forcing the programmer/compiler to structure the code such that the data loaded from memory is used a significant period of time after the load is initiated. This works pretty well on the HPC code that is used on crays, but not at all for the everyday server/workstation code that opterons run. Furthermore, to support that sort of vector unit, you need to have about eight times as much memory bandwidth as an opteron, which means many more pins on the socket, which are very expensive.

I think you're much more likely to see the cray vector processor retooled with lots of hypertransport connections, so it can use an opteron as its scalar unit, and use the same seastar routers that the xt3 uses. On the X1, the scalar unit already runs ahead of the vector unit, so I bet it's not all that important for the scalar unit to be on-die.

Re:excellent

[evilviper]

Saying "I can provide real and reliable sources" without actually doing so is just as dodgy, maybe even more so, than not talking about reliable sources at all.

That makes no sense at all. Anyone with even a passing understanding of the rules of logic would surely disagree with you.

In your own trollish way, you did sort-of ask for some sources, so here's a handful:

http://www.eweek.com/article2/0,,1857009,00.asp
http://www.computerworld.com/hardwaretopics/hardwa re/desktops/story/0,10801,104807,00.html
http://www.theregister.co.uk/2005/09/21/intel_chip set_shortage/
http://www.infoworld.com/article/05/11/04/HNchipse tshortage_1.html
http://news.com.com/2061-10801_3-5850416.html?part =rss&tag=5850416&subj=news
http://www.theregister.co.uk/2005/09/09/intel_chip set_shortage/
http://www.channeltimes.com/channeltimes/jsp/artic le.jsp?article_id=68906&cat_id=883

Re:excellent

[convolvatron]

its a cm5. as to why anyone would choose to build that machine again,
i dont know.

Re:excellent

[AyeRoxor!]

It would make sense only if AMD and Cray improved the tech in a way that would make the Opteron better suited to the server market as well, such as by improving the scalability of the technology.

Perhaps they could license the result to Intel. They don't seem to be doing much lately.

Re:excellent

[Frumious Wombat]

So, just curious, but why don't you put your efforts into crunching your traditional vector processors onto some more affordable piece of silicon, rather than trying to recreate the T3E out of Opterons?

I know chemists who claim there there are still algorithms than don't run as well on modern MPars as they did on mid-90s vector Crays. I know we're not a huge market, but I bet there are some other fields that would rather have a deskside T90, rather than a multi-proc Opteron box.

Re:excellent

[abradsn]

What's the deal with your companies stock? Is it a bad investment or what? It seems that it just keeps going down, causing me to lose more of that investment. Maybe you guys could get some penetration into the smaller sized markets, and increase some revenue so I could feel better about this investment.

Re:excellent

[budgenator]

The best thing they could do for the super-computer market is develope chips with more MIPS per Kcal and more MIPS per watt. It's at the point now where a big super-computer facility or data-center has to call the power company to warn them when they are going to power-up one of those machines. Data centers are pushing 3.8 Kw per square foot now. Its getting hard to get the power cables to the racks physicaly.

Re:excellent

[InvalidError]

AMD's fabs are not that cramped... AMD is still in the process of retooling (upgrading) some of them to 65nm/300mm and it also has some fabs that have been built/upgraded but not up to production yet.

AMD's production is a little tight (probably until the 65nm fabs finish qualification) but deals like Cray's are long-term one-off (more or less) contracts, AMD will not have to churn out those 262k (my guess) Opterons overnight. I would also not be surprised if Cray's order happened to be for those near-future 65nm Opterons.

Re:excellent

[Aardpig]

Nice summary - thanks for the info!

Re:excellent

[pantherace]

AMD usually doesn't retool chip factories, and instead builds complete new ones, every couple of years. This leaves the old ones for flash. (At the moment their flash is struggling, but anyway.)

Intel retools all/most of it's fabs to newer processes. AMD finds it more cost effective to build a new fab every few years. (Indeed, Intel may be less efficient, as I've heard other companies find it more expensive to retool than to build a new one.)

AMD indeed does not have as much capacity as Intel or IBM, they have however contracted with a lot of other fabs, as a contingency measure. If say their sales force were able to sell a lot more Opterons to OEMs, you want a way to make sure you can get those processors made. (At least based on the article, their sales force does seem to be making headway.)

Re:excellent

[darkmeridian]

I guess we have learned what happens when you annoy the AMD-Linux zealots on this website.

Mod away.

Re:excellent

[ryanov]

I'm not 100% certain of this, since I'm working from memory, but I was the admin of a 2 node Origin 200 system for awhile. I'm fairly certain the technology that connected the two was called CrayLink. We also had an Origin 200 in my HS, back in ~1998. Not sure when the Origin 2000 was released -- was it at the same time? If not, it was in the 200 first.

Re:excellent

[ryanov]

WebCT Vista?

http://www.webct.com/products/viewpage?name=produc ts_vista

Re:excellent

[joib]

No they won't! They have no reason to.


Yes, you're probably right that it doesn't make sense for AMD economically. But I want to run numerical codes at more than 5 % peak performance on my cheap Opterons, so I want to believe. ;-)


The vector units that a cray uses aren't like altivec, sse, or other "bolt-on" vector units. The vector unit on a cray (or NEC) is a latency hiding mechanism. It's a method for forcing the programmer/compiler to structure the code such that the data loaded from memory is used a significant period of time after the load is initiated.


Yes, I know. And that's precisely the reason why I'd like to see real vectors instead of the sse/altivec toy ones. Main memory latency is hundreds of cycles, and it's getting worse all the time.

Additionally, from a microarchitecture perspective, vectors have quite a few advantages there too.


This works pretty well on the HPC code that is used on crays, but not at all for the everyday server/workstation code that opterons run.


I'm not sure about that. I guess technical apps vectorize just as well as HPC codes (well perhaps not the UI, but the code that runs the actual simulation or whatever). Heck, even some database code vectorizes nicely (sorting and hash joins).


Furthermore, to support that sort of vector unit, you need to have about eight times as much memory bandwidth as an opteron, which means many more pins on the socket, which are very expensive.


Yes, as I said some Alpha Tarantula like design is probably overkill for the vast majority of the market. My point was that a vector ISA extension with modest execution resources wouldn't need that much die area, and could help make better use of the available bandwidth, whatever that bandwidth is. As you said yourself, the expensive thing is IO. Transistors are cheap by comparison. So not having instructions that allow one to effectively use the available IO resources is a real shame.


I think you're much more likely to see the cray vector processor retooled with lots of hypertransport connections, so it can use an opteron as its scalar unit, and use the same seastar routers that the xt3 uses. On the X1, the scalar unit already runs ahead of the vector unit, so I bet it's not all that important for the scalar unit to be on-die.


Yes, that sounds feasible. IIRC it is something like this that Cray has cooked up for the Cascade project; I.e. a node consists of 8 (or was it 4) scalar processors connected to memory (I guess these could be Opterons or further in the future some kind of Processor-in-memory (PIM) stuff), and a vector unit with its own cache and fast access to the main memory via the scalar cpu:s.

As for the seastar thing, I think you're right that that's what they'll use for inter-node communication. Currently X1(E) uses Numalink licenced from SGI, so they're certainly looking at replacing that with existing in-house tech. BTW, 2H2006 will see the XT4, with the new Opteron sockets with DDR2 memory and the Seastar2 router that provides twice the BW compared to the existing Seastar.

It only makes sense

[heatdeath]

Specialized computing hardware for supercomputers has always seemed like a fiscally bad choice. It'll be good to see what kinds of improvements we can see in research possibilities as supercomputing costs come down from using mass-marketed parts.

Re:It only makes sense

[fgodfrey]

I've never completely understood this argument (yes, I admit, I'm heavily biased). If I want to build a skyscraper, I'm not going to use the "mass market" crane that puts up the roof of a residential house. I'm going to use a specialized crane that's meant for building skyscrapers.

That doesn't mean that there isn't a place for commodity hardware in supercomputing, but to say that there's no room for custom hardware either misses the point. The only thing "off the shelf" about an AMD based Cray is the AMD. The logic board, and, most importantly, the network that interconnects the processors is entirely custom. Not to mention the fact that Cray will still build some entirely custom processors...

By the way - this is hardly the first Cray based on a commodity processor. The T3E and T3D were both Alpha processors, yet nobody calls those machines "commodity".

Re:It only makes sense

[abbyhoffman]

True true, but Jiva really isn't made for IPC like Cray.

Re:It only makes sense

[Seanasy]

The whole point of a supercomputer is that it performs well beyond what commodity systems are capable of. When supercomputers are made entirely of commodity parts, there will be no supercomputers.

Re:It only makes sense

[Jeff DeMaagd]

By the way - this is hardly the first Cray based on a commodity processor. The T3E and T3D were both Alpha processors, yet nobody calls those machines "commodity".

I really don't consider Alpha to be a commodity chip. While Opteron and Athlon64 share a lot of the same designs and even might have the same masks, it doesn't necessarily make it a commodity chip. I suppose either chip were produced in much greater volume than Cray's custom processors.

Re:It only makes sense

[heatdeath]

Darn it, I don't understand you people. I posted this comment as a test, because
  1) it's a shallow unsubstantiated commment
  2) it shows a very low comprehension of what actually goes in to computer systems engineering
  3) it's cliche, and it took 10 seconds to think up.

And, as I suspected, it got modded up to a 5.

Moderating is like voting. It only works if the voters aren't stupid.

Re:It only makes sense

[tzot]

Slashdot readership is not an elite, it's just an assembly of geeks. The up-modding of your article just means that some random readers found it insightful; nothing more, nothing less. It's not a medal or a decoration; judging just by the difference in our /. IDs I would assume you would know by now ;-)

And moderating is voting.

Re:It only makes sense

[heatdeath]

Slashdot readership is not an elite, it's just an assembly of geeks. The up-modding of your article just means that some random readers found it insightful; nothing more, nothing less. It's not a medal or a decoration

That's true. I guess my point, though, is that moderating doesn't actually pull much more insightful comments to the top. You could probably write a bot that would do as good of a job of pulling insightful comments up, just by analyzing word length and grammatical construction.

Re:It only makes sense

[Lost Found]

Convex made the prediction years ago that as the technology powering commodity processors advanced, the market for specialized supercomputer hardware would die off.

Looks like it's happening. The Superdome can either run PA/RISC or Itanium2. Opterons on a Cray system?

And it's not just limited to processors either. I have a bag of sticks of 72-pin DIMMs in the closet (yes, the kind PCs used to use) that came out of a supercomputer. Technology advances in supercomputing will make their way back into the general computing market, where the sales and hence production is much higher.

Besides, the whole "one big machine" theory seems to be slowly dying off as well. We live in the day of high-performance interconnects and 10Gig ethernet.

Going back to your skyscraper analogy, you better well use the mass market crane if it's capable of doing all the same lifting your specialized crane is, because otherwise a competitor might just slip in and outbid you on the job.

Re:Had to be done

[nnnneedles]

Had to be done

No, it didn't.

Can we have a "-1, Catch phrase" option, please? The old jokes are not even remotely funny anymore..srsly.

Hmmm

[Aundy]

I wonder what the governmnet will do with these cheaper, powerful supercomputers?

Re:Hmmm

[turgid]

I wonder what the governmnet will do with these cheaper, powerful supercomputers?

Why, decrypt your 66,000ft high stack of SSH traffic in case you're a would-be terrorist, of course! :-)

Re:Hmmm

[Anonymous Cowpat]

Team slashdot much redouble it's efforts!
The last thing we want is seti@home team 'USGovt' overtaking us!

Re:Hmmm

[NerveGas]

Probably the same things it already does, which in large part is to simulate weather and nuclear reactions.

steve

Re:Hmmm

[Dragoonmac]

Why, with computers running at one petaflop, they should be able to play Doom 3 at medium quality.

Re:But

[fgodfrey]

Yes, actually.

Sign You Invested In The Wrong Supercomputer, #342

[John_Booty]

#34. Your "supercomputing" vendor has an AOL email address.

From the press release...

Contact:
AMD
Teresa Osborne, 512-602-0040
teresa.osborne@amd.com
www.amd.com
or
Cray Inc.
Steve Conway, 651-592-7441
sttico@aol.com
www.cray.com

Sooooo... if I scrape together a few million bucks and buy a computer from these guys, will I still be able to contact my Cray rep once his 500 FREE TRY AOL NOW HOURS have expired?

Re:Sign You Invested In The Wrong Supercomputer, #

[John_Booty]

Signs You're Posting Too Quickly #293893...you type #34 instead of #342. Oops. Sorry!

Sign You Invested In The Wrong Supercomputer, #342
#34. Your "supercomputing" vendor has an AOL email address.

Re:Put on your tinfoil hats...

[B3ryllium]

DARPA created the Internet.

You are using the Internet.

You are part of the conspiracy.

Re:...and now some real numbers

[Vengeance]

Hrmmm. In six months AMD went from 25 systems on the list to 55 systems on the list, and you think Intel is doing well?

Let's extrapolate for a moment, shall we? I'll even do Intel a favor and clamp down on the AMD increases each time. Basically, AMD more than doubled their share of this elite group in six months' time.

Six months from now, they've almost doubled to 100 systems.

Twelve months from now, slowing down and growing only 75%, they've got 175 systems.

Two years in the future, with even more slowing down of their growth, 300 systems on the list are AMD. I wonder whether the preponderance of that growth comes from the current 400-odd Intel machines or from the 73 IBM setups...

Likely? Maybe not. Possible? Yeah, it just might be.

nVidia motherboards

[CCFreak2K]

Hey, maybe the motherboards are nForces too. I bet all the new Crays will have digital 5.1 sound, an important feature for today's supercomputers.

Re:nVidia motherboards

[Rufus211]

Actually at least one of the Crays use the nForce Pro chipset in them.

Re:...and now some real numbers

[hattig]

On November 2004 the list had 84 computers with Itanium 2 processors. In June 2005, the number shrunk to 79.

Now only 46 computers contain Itanium 2 chips according to the latest list, released Monday.

That's quite a collapse. Intel is propping up their high-end systems with volcano-simulator Xeons?

Meanwhile, the number of supercomputers using Advanced Micro Devices' Opteron chips has increased. A total of 55 Opteron-based computers made the list, up from 25 in June. (Opterons were found in just 29 computers on the November 2004 list.)

A near doubling in a year. And that's with AMD's first real server standard processor. HORUS comes out today, that'll put AMD into the 32 and 64 core marketplace. Not bad for a company with 0 server marketshare, nevermind Top500 systems two years ago.

As for the rest of your troll, I think most of the people here are clever enough to see it for what it is.

Re:Put on your tinfoil hats...

[Silverstrike]

This is something I have never understood. What makes you think they care what you do in your day to day life? And if they did have this apparently top-secret information, what horrible and nefarious purpose do you suppose they'll use it for?

Now, its true that in history that the more privacy a government gives its people is often indicative of the quality of life those citizens enjoy. However, that does not imply that the act of collecting information causes the decline in QoL. I'd argue that the inherant distrust in a government that does invasive monitoriing, such as some of the old Soviet regimes, is the problem; the spying is a symptom.

Personally, if they want to know what I ate for dinner for the past week, they're more than welcome to know.

Furthermore, information about my day-to-day life certainly won't empower such people to lock me away in a gulag in the middle of nowhere to rot and die -- they don't need to know what Ihad for dinner to do that. All they need is the power to actually whisk me away in such a manner (just a car and one or two big guys, really) And this power will come seperate from the ability to know what, precisely, it is Ihad for dinner the night before.

Oh, and as an aside, the esteemed minds at MIT have pointed out recently -- that tinfoil hat of yours will only help them discover the content of your dinner plate the previous evening ;-)

The irony of it all.

[Anonymous Coward]

Crays supercomputers were known for their high performance vector operations. These operation have very little to do with PC world except its close cousin - SIMD operations (gaming, graphics). Now the fact that AMD tops cray (at least on commercial merits) is like having an AMD instruction set adopted by Intel (oh, wait).
More ironic is the fact that the compiler that will be used for those supercomputers is probably the PathScale variant of Open64 - SGI's compiler that was released as open source after it was retargetted to the Itanic architecture.
I might have some misconceptions, careful readers, please fill-in the blanks.

Re:The irony of it all.

[Shoeler]

Aren't you missing the point though? It's not that the AMD systems are prevailing on the merits of their amazing vector math (they aren't) - it's that they do a PRETTY good job of both vector and scalar math, but at the prices you can get them, your cost per computation is SIGNIFICANTLY lower then it would be with one of the massive vector systems.

The research / development arm of the organization I work for just got a 4000+ CPU XT3. Last I checked, they planned on using the PGI compiler for most stuff.

Re:The irony of it all.

[convolvatron]

disclaimer: i worked at cray on the xt3

its not ironic at all. its a question of resources and volume.
cray has a few very bright people (still, sort of). they are
essentially a us government lab. they do a bad job, but its insane
to think that 100 people can build and maintain several different
supercomputer architectures.

a $300 opteron is almost always more effective than a $60000 X1
processor. they have alot of bright people too, and alot more of
them.

the only reason that cray still exists is support for parallelism
and the provision of high memory bandwidth systems. but even that
niche is being eroded pretty severely. the xt3 communications chip
runs at 3.5GB/s in each direction. it costs about $250 for cray to
have each of them made. for the same $250 i can buy a mellanox nic
that runs at half the speed

its no suprise that cray is using opterons. they actually got lucky
by committing to amd early and having it turn out so well.

the real question is whether there is any more room for a cray at
all. the commodity world moves so quickly. the xd machines (which
they purchased) are really their best asset, but it still hard to
justify that kind of margin for what is essentially a well
constructed cluster.

Opteron is not NexGen's tech

[tomcres]

K6 technology was acquired and modified by AMD. The K7 and K8 were designed by AMD. True, many of the engineers on the K7 and K8 teams were probably ex-NexGen since AMD acquired that company, but so what? They are truly AMD innovations. At least they didn't sink all of their research into the Itanic!

Re:Opteron is not NexGen's tech

[sgt scrub]

Actually the K7 was designed with most of the DEC technology. I'm not knocking AMD though. I've used them since the 386 days. I think they have surpassed the Alpha --me ducks-- in speed and efficiency, though I still wouldn't give up my 433a withought a fist fight.

Their bus arch and chipset tech is the most interesting. (if someone has proof that AMD didn't design this it better be solid). This attention from Cray, and the super computer people in general, is due more to this success. AMD has the best design and it shows. It is one thing to buy schetches of a something and another to make it fly this good.

More to the point regarding Cray is their XD1. THAT is a cool machine! I was looking around at different FPGA stuff and almost shorted my keyboard with drool. Damn, I wish I was rich. -sniffle-

Re:Opteron is not NexGen's tech

[mikefe]

Who says they can't use segmented memory? I'm sure the memory isn't connected by one interconnect, but spread out among the various processors in the system. If the memory is X number of hops away, then it is not accessable (that would be a practical constraint anyway for performance reasons).

Also, why can't AMD make processors with more than 40 bits of addressable memory? They don't have 64bit addressable memory because of cost considerations, though it isn't that hard to expand the limit, relatively speeking.

Re:Opteron is not NexGen's tech

[raxx7]

Simple: they're not shared memory systems.
XD1 is built on 2 socket smp blocks, with Opteron 2XX series.
XT3 (RedStorm) is built on single socket blocks, with Opteron 1XX series.

Finally....

[Prince Vegeta SSJ4]

A Computer capable of running Duke Nukem Forever....oh wait...

my experience

[pigwiggle]

I've been running on an XT3 now off and on; when it's stable it's a workhorse. Anyway, I'm not up on all the given bench marks, etc. But, in my experience (molecular dynamics) with my homebrewed code, an opteron cluster absolutely wastes anything put together with intel or IBM.

Re:my experience

[convolvatron]

when is it stable?

The Chipset is the key.

[mosel-saar-ruwer]

Specialized computing hardware for supercomputers has always seemed like a fiscally bad choice. It'll be good to see what kinds of improvements we can see in research possibilities as supercomputing costs come down from using mass-marketed parts.

Cray likes to build classical vector-driven machines. In that space, you can't rely on some external kludge like Myrinet for your communications; instead, your value-add is in the chipsets that get all those CPUs talking to one another [and to the memory subsystem].

In one of Cray's previous incarnations, they once possessed a chipset/backplane tech for the Sparc processor that Sun purchased off of Silicon Graphics for a song and a dance, and immediately turned into the insanely profitable Sunfire series. The big question here is whether this new agreement requires Cray to share their chipset/backplane tech with AMD [in which case some of it might filter its way back down to the level where mere plebians like us would be able to afford it].

Re:The Chipset is the key.

[flaming-opus]

Of course, cray didn't invent the 6400 either. They bought out a competitor in bankruptcy. Sun, at that time, wasn't quite ready to go head-to-head with s/390s, which is essentially what they did when they finally got around to selling the s10000 (same computer, new name). Chipset is one key, but software is an even bigger key.

It had to be said...

[rubberbando]

Finally! Something that can run Windows Vista at a descent speed... :-P

Re:It had to be said...

[SquadBoy]

I don't know. Descent was pretty damn fast.

Re:It had to be said...

[taskforce]

Windows Vista at descent speed?

Assuming you mean on Earth, so we take G to be 9.8N and a copy of Windows Vista (boxed of course) would weigh 0.5KG, dropping it from a height of 1M: M*G*H would tell us that 0.5*9.8*1= 4.9ms^-1

Re:...and now some real numbers

[Anonymous Coward]

Wow, seems Intel is still 6x AMD market share.

Wow, seems AMD *doubled* it's share of spots in the Top 500 list in *six months*. I bet Intel is ticked, and worried...this is very good PR for AMD.

Go AMD! Milk that NexGen core for all its worth, too bad you didn't invent it, you just bought it.

LOL! Intel fanboys don't have anything real to say these days, they have to resort to cheap ad-hominems. Don't worry, I'm sure someday Intel will come out with competitive chips again. Pretty sure, anyhow.

And as to AMD "just buying it", how would that relate to Intel getting so much Alpha technology and talent from it's deals with HP/Compaq/DEC?

It would be nice if you would start innovating one of these days.

Yeah, if AMD can produce better processors than Intel without innovating, just imagine what'll happen when it does innovate...! =)

Re:Put on your tinfoil hats...

[Mister Yoan]

So privacy isn't a right, it's a privilege?

Behold!

[katana]

I give you...the Crapteron!

Re:Behold!

[sharkey]

It comes with Vista pre-installed then?

Does this mean...

[__aaclcg7560]

That prices on the older Clay computers will drop? Holy flipflops! Now I'll have something to put into my empty warehouse building. :P

Re:Does this mean...

[masklinn]

Yep, as well as the price on Mud computers.

Re:Sign You Invested In The Wrong Supercomputer, #

[flaming-opus]

Cray is a small company.
They probably hire an outside comunications firm to do public relations.

Shall we play a game?

[akgw]

How about Global Thermonuclear war?

Not to be a jerk...but this is old news....

[jamesgomez]

dated from June 16, 2005

Check out the article here...
http://www.hypertransport.org/consortium/cons_pres srelease.cfm?RecordID=79/

Re:Sign You Invested In The Wrong Supercomputer, #

[Nethead]

No doubt. It was almost a month of walking past them on the way to work before I even noticed them. Very low key space. Nice area of town to work in though... endless lunch choices!

Re:...and now some real numbers

[Anonymous Coward]

Go AMD! Milk that NexGen core for all its worth, too bad you didn't invent it, you just bought it. It would be nice if you would start innovating one of these days.

Go Intel! Milk that RISC design for all it's worth, too bad you ripped the idea off from ARM and stuck three metric tons of x86 mentally-backwards-compatibility cruft on it. It'd be nice if you start innovating one of these days, instead of ripping off things like AMD's invention, x86-64.

That's just the technical side. I don't need to go in much detail about how Intel fucked up its marketing by pushing MHz (maybe because they were bitter that AMD had 1GHz first?)

Sun..

[Bulmakau]

I was just about to buy 40 sun machines, based on AMD. Maybe I should wait for Cray to come out with their product? Anyone knows the estimated retail price that machine is expected to hit the market with? ;)

Re:Sun..

[iamlucky13]

I know you're trying hard to make it look this is work-related, but for goodness sake, don't make major purchasing decisions based on what you read on Slashdot!

Re:Sun..

[Bishop]

Insider sources say that the smaller 1024 node model will have a MSRP of 10 MegaBucks. Ofcourse shipping, installation, and remodelling the server room are extra.

Cray already offers the Opteron Based XD1. The smallest 12 processor model is estimated to cost $100k USD.

It should have been said that ...

[DrSkwid]

Finally! Something that can run Windows Vista at a decent speed... :-P

unless you did mean it goes down real fast !

Compilers !! ???

[Anonymous Coward]

But what compilers are people using with the Opteron to get its best speed ? We have been looking for good compilers (Fortran, C, C++) for numbercrunching on AMD-64 and have been dissappointed so far. Our preferred compilers are Intel, but they have been modified to crap out on AMD chips.

Re:Compilers !! ???

[cimetmc]

Have you tried Pathscale compilers http://www.pathscale.com/ekopath.html? They seem to give quite decent performance on AMD64 chips.

Re:Compilers !! ???

[hackstraw]

Have you tried Pathscale compilers http://www.pathscale.com/ekopath.html? They seem to give quite decent performance on AMD64 chips.

I haven't used their compiler, but this looks fscking sick: http://www.pathscale.com/infinipath.php

AMD has come a very long way the past few years. If they had Intel's fabrication skills they would be almost unstoppable. This HTX stuff simply looks like the bomb.

Re:Sign You Invested In The Wrong Supercomputer, #

[wpiman]

I have one too. No idea who pays for it-- but it is great to have for a publicized email address because
A. I don't use the account very often except when I subscribe to a webpage or something.
B. I get a ton of junkmail there anyways.
Saves my work, yahoo, gmail, and home emails from getting flooded.

The answer is quite simple

[drgonzo59]

If they think they'll make more money in the long run, all things considered, they'll do it.

Re:Had to be done

[P3NIS_CLEAVER]

In Korea old jokes aren't funny anymore.

Re:Put on your tinfoil hats...

[Neoprofin]

To be fair, AMD is proud of helping spy on it's citizens more efficiently and we can all be proud of increased productivity. If that wasn't funny maybe it was insightful?

Get your formula's right!!

[Wierdy1024]

Correct me if I'm wrong, but if you want the speed it will hit the floor it's worked out by:

V^2=u^2+2as

so

v^2=0+2*9.8*1

v=4.4 m/s

PS. where is the square root button on windows calculator when on scientftic view??!!

Re:Get your formula's right!!

[yuriismaster]

sqrt(x) = x^1/2

sqrt(19.6) = 19.6 then x^y then .5

Re:Get your formula's right!!

[cannon fodder 0109]

PS. where is the square root button on windows calculator when on scientftic view??!!

There isn't a specific square root button.

You have to hit the "Inv" (inverse) check-box and then the "x^2".

Alternativly you could use the "x^y" button and then 0.5

Re:Get your formula's right!!

[Wierdy1024]

thats what I ended up doing.

Solaris?

[turgid]

Solaris 10 works really well on large NUMA boxes (better than Linux), and it has an Opteron port...

Re:Solaris?

[Zemplar]

Solaris 10 does not have an Operton port. Solaris 10 is available for Sparc and x86. Period. x86 just also happens to include Opterons.

Re:Solaris?

[turgid]

Solaris 10 does not have an Operton port. Solaris 10 is available for Sparc and x86. Period. x86 just also happens to include Opterons.

Wrong! There was a specific port of Solaris 10 to Opteron (AMD64). I know some of the very nice people who developed it and I got to do some development on it while it was under development too.

The NUMA code for Opteron and the 64-bit addressing is all in there. Opteron (AMD64) has an extra layer of page tables that aren't there on regular x86, and some subtle but important differences in the ways it handles the TLBs.

Intel EM64T is Opteron-compatible, but they've been trying to play that down becasue it was late, slow, incomplete and competed directly with the itanic.

Re:Solaris?

[Zemplar]

I've never heard of that, but I stand corrected.

You should understand that I was just trying to correct anyone from thinking that since Sun has adopted and marketed the Opterons that there was no reason to exclude Solairs for other x86 platforms.

Re:Solaris?

[turgid]

You should understand that I was just trying to correct anyone from thinking that since Sun has adopted and marketed the Opterons that there was no reason to exclude Solairs for other x86 platforms.

Indeed. Opteron (and hence Athlon 64) are perfectly backwards compatible with earlier x86 chips, that is they pretend to be a 286 when powered up, and switch into 386 protected mode just like any other legacy x86 processor (including Pentium 4).

The 64-bit goodness comes from the extenstions to the architecture which have to be enabled and handled in the OS kernel.

You're 100% correct in saying that Solaris 10 is still also developed for and supported on 32-bit x86 processors as well, and the 32-bit kernel and environment will also run perfectly on AMD64 processors.

Solaris rules (so does Linux).

Jiva

[garrettkelly]

I'm part of the team that worked on JIVA. This is really comparing apples and oranges. It's meant for a different market than Cray is targeting. Cray is excellent for IPC-intensive tasks like CFD. JIVA is meant for naturally parallel applications and is much simpler to program.

Ha!

[turgid]

This is really just a marketing play on AMD's part.

Sounds like sour grapes from an itanic customer (or an intel or SGI staffer) :-)

Re:Ha!

[InvalidError]

Opteron making major inroads into the HPC market is bad for IBM's Power4/5 and lesser players like NEC too.

Re:Put on your tinfoil hats...

[HAMgeek]

DARPA created the Internet.

You mean it wasn't Al Gore?

"During my service in the United States Congress, I took the initiative in creating the Internet." --Al Gore from an interview with Wolf Blitzer on CNN's "Late Edition" program on 9 March 1999.

The new juncture will be called....CRAPTERONS

[digitaldc]

Just kidding!

Re:The new juncture will be called....CRAPTERONS

[evilviper]

There has never been a comment on /. more deserving of a "-1 Redundant" mod.

While we're on the subject of burnt karma....

[HotNeedleOfInquiry]

Imagine what a Beowulf cluster of these Crays could do...

Re:...and now some real numbers

[DCstewieG]

Did you hear?! Intel pays a ton of engineers to make cores...and they put their own name on it! Too bad Intel can't invent something itself.

AMD

[certel]

Wow. Go AMD. They're definitely have a huge impact on the market. Intel needs to get back in gear.

Re:why not use the ultrasparc T1?

[NerveGas]

They have 8 cores!

So? How much memory bandwidth do they have? Not I/O bandwidth, but memory bandwidth. I highly doubt that they have as much bandwidth PER CORE as the Opterons do, and in big applications, memory bandwidth can be a very important factor.

You cant build an enterprise machine without Ultrasparc (or Power4 or PA_RISC) CPUs.

    I guess that Cray thinks differently.

Re:Put on your tinfoil hats...

[smittyoneeach]

Yes, oh flogger of dead horses: yes.

HOT HOT

[betims]

I think there will be very hot :D

Re:Put on your tinfoil hats...

[networkBoy]

your sig seems aptly appropriate for this comment ;-)
-nB

Makes me wonder ...

[cpu_fusion]

The continued big-name backing of AMD (e.g. Sun, Cray) makes me wonder how sweet a deal Apple must have gotten to go with Intel over AMD. :)

Re:Makes me wonder ...

[tomstdenis]

Intel and Apple are like those kids in school who always got to go on ski trips with their year passes and high end skis, etc. They think they're all important because things are good. Then when they get into the real world they get eating alive.

Apple as far as I can see has a bit of a disconnect with "reality". They're not customer-oriented [e.g. shitty dead pixel policies, really costly vendor-locked in gear, etc]. It was always "oh that's apple gear only because the quality is higher" yet all the folk I talk to say "just get a wintel laptop and be done with". And it isn't like the PPC isn't a nice processor [the G4 actually looks fairly sweet]. It's just they're a bunch of pricks.

Apple went with Intel for appearance sake only. A Turion-based laptop would have done them just fine [or hell just invest and extend the PPC line]. I mean the G5 was a bit extreme, no need for that in a laptop [though it is cool]. A 32-bit laptop is mighty fine given that you don't normally run multi-GB database engines or whatever on it.

Tom

Re:Makes me wonder ...

[masklinn]

Apple as far as I can see has a bit of a disconnect with "reality". They're not customer-oriented [e.g. shitty dead pixel policies, really costly vendor-locked in gear, etc]. It was always "oh that's apple gear only because the quality is higher" yet all the folk I talk to say "just get a wintel laptop and be done with". And it isn't like the PPC isn't a nice processor [the G4 actually looks fairly sweet]. It's just they're a bunch of pricks.

Maybe Apple's computer "just work" and look cool to boot (and not the "hey let's put some more shinies, neons and LEDs in this I'm not sure 3.5MW worth of them is enough" cool) ?

They are pricks indeed, and suckers, but they computers work and OSX is a quite good piece of software.

In fact, quite a few people are currently shifting from Wintel to Apple 'puters...

Apple went with Intel for appearance sake only. A Turion-based laptop would have done them just fine [or hell just invest and extend the PPC line]. I mean the G5 was a bit extreme, no need for that in a laptop [though it is cool]. A 32-bit laptop is mighty fine given that you don't normally run multi-GB database engines or whatever on it.

Flash news: Current AMD Turions are, in fact, 64 bits (which *may* be why their name really is AMD Turion(TM) 64)

Re:Makes me wonder ...

[tomstdenis]

Flash news: Current AMD Turions are, in fact, 64 bits (which *may* be why their name really is AMD Turion(TM) 64)

The 32-bit comment was about PPC not Turion....But yes I did word that wrong and it was misleading. Sorry.

The biggest flaw I see with most incarnations of PPC [like the G4] is they shoot themselves in the foot with bandwidth. 133Mhz FSB? What's this the mid 90s? Put a good ol' dual-channel DDR400 [or DDR-II-533] on the front of it and be done with.

I'm sure with a slightly longer pipeline [iirc the ALU in the G4 has 7 stages] and an integrated dual-channel memory controller you could kick out 32-bit PPCs that would give the rest a run for their money. I'm not suggesting this is trivial or cheap. Just suggesting it's a worthy project.

Keep in mind the PPC has some clear advantages over x86. The instruction set is RISC so there is no need for advanced decoders [e.g. to make macro-ops and the like], there are 32 GPRs and FPRs available making code run off registers a lot easier, etc.

Crank a G4 like processor up to 2Ghz or so and you'd see cool shit happen. :-)

Tom

Re:Makes me wonder ...

[Watts Martin]

My impression is that Intel has been irritated that they come up with new technologies that the PC world isn't particularly interested in. Despite the "Wintel" moniker, Intel has had a love-hate relationship with Microsoft; back in the BeOS days, Intel offered engineering assistance (and I suspect financial support) directly to Be to get the OS ported. And today, Apple doesn't have 20 years of backward-compatibility hardware baggage to deal with, so they have the potential to be a showcase partner in a way that the Dells of the world just aren't.

As for why not AMD, though, Intel has placed a much higher focus than AMD has on very low-voltage chips, and from what I've heard, that's what ultimately gave them the Apple nod. Arguments about production capacity aside, AMD doesn't have the R&D resources of Intel, and they have to pick their battlegrounds carefully. They've picked them wisely, but as of right now, they don't have anything competing with chips like Intel's low-power, dual-core lineup for 2006. If that's remedied in 2007, I'm sure Apple would have no problem revisiting it, but right now, AMD just doesn't have the chips they want.

(As for the production capacity arguments, I've seen people here point out that Intel has had problems meeting their demand recently and AMD hasn't. While this is true, it's important to keep in mind that Intel's overall demand is still over five times that of AMD's, and the gap in notebooks -- the segment where Intel's production capacity fell behind demand temporarily -- favors Intel by an even greater margin.)

Re:Makes me wonder ...

[be-fan]

You can't crank a G4 up to 2GHz, because of the short pipeline. You could do like IBM, double the pipeline length (to 16 stages), and crank it up to 2GHz, but then you have a G5. The G5 is a decent chip, but out of the trio of Pentium-M, Opteron, G5, it's the slowest. It also is severely hampered by GCC. IBM's compiler can make a 2.5 GHz G5 perform comparably to a 2.2GHz Opteron in integer, and a 2.8 GHz Opteron in FP. Using GCC, it's more like a 1.8 GHz Opteron in integer and a 2.0 GHz Opteron in FP.

Re:Makes me wonder ...

[ChunderDownunder]

Crank a G4 like processor up to 2Ghz or so and you'd see cool shit happen. :-)

Well, maybe not 2Ghz overnight, but there's life in the G4 yet: dual core...

I'd rather have one of these powering a mac-mini than a celeron-m. Such is life...

Re:Makes me wonder ...

[wild_berry]

I admire IBM's commitment to Free/Libre Software that their own personal compiler does worse than the GNU version. Their knowledge of their own hardware and their not sharing that part of their technology with GCC is their own choice and one which makes them look bad.

AMD makes more than chips!

[scoobrs]

The parent poster needs to be reminded that a large chip manufacturer like Intel, IBM, and AMD makes much more than CPUs! They play a fundamental role in the design and system architecture of the machines built out of their chips. Interfaces like Hypertransport, PCI Express, and DDR are the work of these chip giants. To claim that changing the fundamental design of the CPUs has anything to do with the interaction of a supercomputer company and AMD is naive. Far more likely are changes in Hypertransport, interfaces to memory, or other bus-level projects that are more useful to a supercomputer vendor looking for the best possible overall system bandwidth anyway.

When does this translate to bank?

[Qrlx]

AMD chips outperform Intel, for less money, using less power, for something like five years now.

Shouldn't AMD stock be doing better? If you bought 5 years ago, AMD is flat, but Intel is up like 50%.

Re:When does this translate to bank?

[tomstdenis]

Is you confused or something? This is what theys call in America "modern capitalisms".

And besides you're supposed to use retarded circular and self-defeating logic like

"AMD doesn't have the big customers because they can't afford to upgrade their supply chain to have the big people as customers." -- Average retarded suit.

Tom

Re:When does this translate to bank?

[jimicus]

Stock prices seldom reflect what the company is actually doing, and are more often driven by either the market in general or mindless "analysts" who wouldn't know what AMD does if a chip were to bite them in the bum.

Re:When does this translate to bank?

[masklinn]

Since when does stock value reflect a company's technical prowess of potential?

Re:When does this translate to bank?

[abradsn]

Stocks are a futures market, so (sometimes) this is exactly what they represent.

Re:When does this translate to bank?

[masklinn]

That may be what they're *supposed* to represent, but that sure ain't what they *do* represent.

Re:When does this translate to bank?

[freemancomputer]

What it comes out to is that intel has been around far longer then AMD and it the early years intel was able to start those love hate friendships that everyone talks about. AMD cators to the computer geek end users and we all know how the advrige geek is, broke or they hord all there moeny. I think that if, big if, AMD wins there law suite aganst intel that there stock will go up along with there RnD budget and with that they will make better cheeper chips that will give them a better market share. I dont think that it will ever be 50-50, and to be honsit I hope that it never is. Being that AMD has to make up ground in the market place gives them the push to make better chips. If they over take intel I dont think intel will make the push to make things better to over take AMD and at that point I think that things will slow way down.
IMHO

Re:why not use the ultrasparc T1?

[dk.r*nger]

So? How much memory bandwidth do they have? Not I/O bandwidth, but memory bandwidth. I highly doubt that they have as much bandwidth PER CORE as the Opterons do, and in big applications, memory bandwidth can be a very important factor.

The Niagara has four 144-bit interfaces.
The Opterons (Both single- and dualcores) has two 72 bit interfaces.

The Crays are not just about memory bandwidth, but also alot about an efficent interconnect.

OT: It only makes sense

[Bishop]

Moderating is like voting. It only works if the voters aren't stupid.

Signal11 proved this years ago. He posted many posts conforming to your points 1 and 3. Especially 3. IIRC many of his posts were longer which gave the post an air of insight, when in reality the post was just fluff.

Why?

[nurb432]

Are there not much better chips out there? ( like PPC, MIPS, SPARC, totally custom ASIC's.....)

I guess being a commodity chip helps for supply issues, but when you are building machines that are this expensive, is that really a deciding factor?

Re:Put on your tinfoil hats...

[B3ryllium]

What can I say? I'm on the Jazz, man ...

Re:why not use the ultrasparc T1?

[turgid]

Because they don't do floating-point in hardware, or at least not to any useful level of performance.

The 8-core Niagara (T1) has 1 floating-point execution unit on only 1 of the 8 cores. Buy a 6- or 4-core Niagara, and do you get a floating-point execution unit at all?

On Niagara (aka UltraSPARC T1) floating-point will mostly be accomplished with software emulation of the SPARC V9 FP instructions.

That's why you wouldn't use Niagara for supercomputing. Web serving, yes, computational fluid dynamics or numerical general relativity, no.

The Irony of Marketing

[scoobrs]

The parent poster posts well for one ignorant of the simplest precepts of marketing. The first things a marketer learns is he must segment a market and only compete in the segments or niches in which competition is profitable. Cray isn't competing directly against clusters because clusters don't have the bandwidth necessary for the sorts of problems Crays are aimed at and Crays tend to be overkill for the problems clusters are aimed at. Cray doesn't seek out customers $.5M for that reason. Anyone who actually uses the supercomputers to solve problems knows that a 50% difference in interconnect speed per single link could mean a 90+% slowdown on a large system using a large program with high overhead. Plain old clusters aren't targetted against Crays, except by some communities that don't buy supercomputers for supercomputer problems anyway, like most Slashdot users. In the supercomputer world, MTTI is everything! That means mean time to interrupt. A bad memory module or a CPU fan blowing out on your single CPU might happen every 3 years on average, but multiply these sorts of problems by 10,000 CPUs on a supercomputer and your cluster will never get any useful work done before something goes out and it crashes. Disclaimer: I worked on the X1/X1e, which is still faster than any other chip on select problems which vectorize well. I agree that the AMD partnership was and continues to be an excellent decision, but it only says that AMD does SCALAR performance better, not everything!

Re:The Irony of Marketing

[convolvatron]

are you seriously trying to claim that cray occupies a
position of marketing excellence?

from a business point of view its completely whacked. cray spends
4-5 years of time to build a machine, just to sell a very small
few of them, throw almost all of the technology away and start
over again.

and are you really trying to claim that any cray machine built in
the last 10 years has particularily good mtti rates? the sv2
really was basically unsuable for the first two years after it
was shipped. its still kind of a dubious proposition.

Re:The Irony of Marketing

[scoobrs]

cray spends 4-5 years of time to build a machine, just to sell a very small few of them, throw almost all of the technology away and start over again.

That's exactly the problem Cray has had competing with microprocessors. It's not that microprocessors are technically superior to custom vectors for that market. They're not! It's that Cray's market segment is too small to support the fast development cycles and huge research budgets of chip companies like AMD and Intel. By latching onto AMD's products for some of the work, Cray can avoid spending so much on a development cycle. I'm not trying to claim that the marketing has been perfect, but that they're at least considering the problem unlike kneejerk clusterheads. For the size of some Crays sold, the MTTI has been quite respectable.

Re:The Irony of Marketing

[manno]

MTTI... a little help

Good designs?

[tzot]

Don't worry, I'm sure someday Intel will come out with competitive chips again.

Pentium M is a great chip. Far better than the Pentium 4, anyway.

I'm a fan of AMD for years now, but I appreciate a lot the performance of the Pentium M in my laptop. So now I wouldn't mind using a cluster running on Pentium M server chips...

Finally :)

[Nonillion]

I can afford that 4-way Crayteron workstation I have been dreaming about..

XD1 has 6 FPGAs

[jaywee]

Yeah, that's precisely why XD1 has not only 12 opterons, but 6 FPGAs too. Which can be used for various kinds of specialised stuff.

Re:Sign You Invested In The Wrong Supercomputer, #

[MauricioC]

The question is... will the Cray actually come with AOL CDs?

Get your apostrophes right!

[doodlelogic]

Did you mean get your formula is right? Or did you mean the right belonging to the formula? No, so no apostrophe. If you're going to correct someone, do it right.

Re:Sign You Invested In The Wrong Supercomputer, #

[Ilgaz]

There are very advanced coders using @aol.com account which surprised me a lot once.

I think it has something to do with how many years guy is on internet? Or massive space aol offers? (before gmail)

Re:Put on your tinfoil hats...

[t35t0r]

who has tin foil anyways? You mean aluminum foil?

Re:Put on your tinfoil hats...

[B3ryllium]

Hah, you jest, but I have already posted on the subject. I grok the conspiracy, man, and it's SCARY.

FAB 36

[charnov]

Well, FAB 36 is open and doing preliminary production runs on 300mm wafers at 65nm. It is supposed to be at 13000 wafers per month by 1Q 06 (or sooner) and 20000 wafers per month by 4Q 06. FAB is supposedly being retasked to do 300mm at 90nm and chipsets for both AMD and its partners.

Those quantities should get AMD to about 25% of the worlwide supply of x86. They are at 19% right now for comparisons sake.

this is old news

[sl4shd0rk]

Cray has been using Opterons for quite some time now. They chose the Opteron because that is what Hypertransport is for - hooking together several processors. Cray has a rack that holds 12 blade servers with each blade housing 6 Opterons (iirc), 2 Opterons per daughterboard. A flavor of Linux runs at the core. blah and blah

Re:Had to be done

[sr180]

In Soviet Russia, the Jokes turn YOU old.

spam protection

[Barbarian]

This is obviously for spam protection...

Last release of Open64: 2003-03-21

[Paul Crowley]

Open64 appears to be moribund: it has been over two and a half years since they last made a release...

http://open64.sourceforge.net/news.html

Just imagine

[zde]

New, huge, shiny, 64-bit, fully virtualizabe supercomputer cluster... booting... in 16-bit real mode.