IA64 and emulator performance
|
From: Casper H.S. Dik on 23 Oct 2006 05:35 jgd(a)cix.co.uk (John Dallman) writes: >You aren't the only one. The claim may be based in the fact that IA-64 >processors are on general sale to anyone who wants them. However, POWER4 >was formerly used by Apple, as the "G5", and the SPARC architecture >definitions were generally available, as were chips, if you wanted them. >I tend to read the "IA-64 is not proprietary" claim as that special >category of truth known as "marketing". You can make a SPARC chip (and also a PowerPC chip?) without having to pay licensing to anybody. That is not the case for the proprietary IA-64 chips which are single source and you're not allowed to clone them. Intel's reasoning must be something like this: IA-32 is "industry standard" and non-proprietary (it has indeed several different suppliers), therefor so is IA-64 (because there's only 32 digit difference and we make that one too). It's probably the most proprietary CPU chip around. >The last two groups don't tend to believe that there is a real risk of >Itanium disappearing, largely because they believe the statements to >that effect from HP and Intel executives. The problem is that (a) Intel >is now in complete control, because HP sold their IA-64 design teams and >intellectual property to Intel, so HP's controls are purely contractual >(b) Intel will, in the end, do whatever it takes to survive, and some >point, they'll get fed up of loosing money on Itanium. It's probably the 10 billion dollar question: how much do they have to pay HP if they stop shipping Itanium? Death by attrition seems more likely; make it less and less compelling until HP's customers stop buying them and Intel can say to HP, you don't buy anough, we're stopping. Casper -- Expressed in this posting are my opinions. They are in no way related to opinions held by my employer, Sun Microsystems. Statements on Sun products included here are not gospel and may be fiction rather than truth.
From: Bernd Paysan on 22 Oct 2006 18:29 Greg Lindahl wrote: > In article <memo.20061021155512.2712B(a)jgd.compulink.co.uk>, > John Dallman <jgd(a)cix.co.uk> wrote: > >>Absolutely. It's serving as a demonstration of the "supercomputing >>fallacy", the belief that almost everyone has problems that are amenable >>to the techniques of the category of machines that have been known as >>"supercomputers". > > Perhaps the "supercomputing fallacy" is that people have no clue what > chip features make good supercomputers? Certainly the Itanium is > missing the most imortant one, "significantly better performance than > the rest", and the next important one, "great price performance". Yes, reminds me on a report from Leibniz Rechenzentrum (here in Munich) which got an Itanic-based supercomputer. They are still waiting for the processor upgrade, while the price/perfomance ratio of their machine is about four times worse than of a comparable Opteron-based system - not the one they have now, but the one *after* the promised upgrade. The upgrade is to what they've actually ordered. -- Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
From: Eugene Miya on 23 Oct 2006 17:22 In article <memo.20061021155512.2712B(a)jgd.compulink.co.uk>, John Dallman <jgd(a)cix.co.uk> wrote: >>>Absolutely. It's serving as a demonstration of the "supercomputing >>>fallacy", the belief that almost everyone has problems that are amenable >>>to the techniques of the category of machines that have been known as >>>"supercomputers". That's the belief. The reality is that as performance goes up, the actual number of cited grand challenge problems goes down. It's easily counted. Greg Lindahl wrote: >> Perhaps the "supercomputing fallacy" is that people have no clue what >> chip features make good supercomputers? Certainly the Itanium is >> missing the most important one, "significantly better performance than >> the rest", and the next important one, "great price performance". True and true. The keyword is "significantly". Transputers would be around more were that true. In article <eu7t04-k51.ln1(a)vimes.paysan.nom>, Bernd Paysan <bernd.paysan(a)gmx.de> wrote: >"If you want it done right, you have to do it yourself" Start with sand. 8^) --
From: Eugene Miya on 23 Oct 2006 18:24 In article <memo.20061021155512.2712B(a)jgd.compulink.co.uk>, John Dallman <jgd(a)cix.co.uk> wrote: >|> >Absolutely. It's serving as a demonstration of the "supercomputing >|> >fallacy", the belief that almost everyone has problems that are amenable >|> >to the techniques of the category of machines that have been known as >|> >"supercomputers". In article <453be339$1(a)news.meer.net>, lindahl(a)pbm.com (Greg Lindahl) writes: >|> Perhaps the "supercomputing fallacy" is that people have no clue what >|> chip features make good supercomputers? ... >|> the next important one, "great price performance". In article <ehi05m$liq$1(a)gemini.csx.cam.ac.uk>, Nick Maclaren <nmm1(a)cus.cam.ac.uk> wrote: >It did have "significantly better performance than the rest" - for a >while, and for the very small number of Itanium-friendly applications! > >Unfortunately, as with many other specialised architectures, there >weren't enough of them to make a market. And I agree that, even for >them, its price/performance was poor. > >|> The action in supercomputing is mostly in the interconnect these days. Well interconnect was where some people thought it was since the 70s. This, of course, irks the software people and why PVM came out. Some applications can do all processing locally, others require data at some distance. >|> The attack of the killer micros is sooo yesterday. Somewhat. >Indeed. Provided that you include memory access as being part of the >interconnect problem, which it is nowadays. Oh, depends how one looks at memory. Well one thing for certain: a solution is not coming out of the UK. --
From: Eugene Miya on 23 Oct 2006 18:51
In article <memo.20061021155512.2712B(a)jgd.compulink.co.uk>, John Dallman <jgd(a)cix.co.uk> wrote: >In article <1161384017.296321.183600(a)k70g2000cwa.googlegroups.com>, >jsavard(a)ecn.ab.ca () wrote: >> it seems to me that the Itanium is the closest thing to a mass-market >> supercomputer chip there is. Supercomputers are complete and ideally balances systems. Chips are merely components in a system. A slew of poorly balanced systems show that merely throwing the best/fastest components together does not guarantee a fast system. >Absolutely. It's serving as a demonstration of the "supercomputing >fallacy", the belief that almost everyone has problems that are amenable >to the techniques of the category of machines that have been known as >"supercomputers". Some people have Kryptonite applications. Architectures are known to wilt under their cycles. >Such problems have historically been large-scale simulations, where the >market for solutions was willing to go to great lengths to coerce their >problems into forms suitable for special-purpose hardware, because they >needed solutions really badly. Well that market was one where the guys who had the money called the design. It was a smaller market infrastructure at that time. Simulation is merely the most public of those Krypton-based applications. >people who have problems that they can solve on general-purpose >computers, slowly, and don't feel it worth ("find it extremely hard") >coercing into a form highly suitable for IA-64. I work for one of them, .... >> Which may not be saying much, of course... but it definitely has >> at least a few features the Pentium lacks, and it must be justifying >> its price with some degree of increased performance over common >> garden Pentium IV chips? > >Oh, yes. But the justifications are mainly psychological, not technical. Economic and political. >The people who are buying Itanium systems fall into several distinct >groups: > >* People with supercomputing problems that fit really well onto Itanium, > who are prepared for the idea that Itanium may go away in a few years. > That seems to be OK for "real supercomputing" people, who are used to > reworking their codes for different hardware. Some people get into supercmputing not knowing that they have to rework codes if they want performance. >* People who have taken on HP's line that it's the natural successor > to PA-RISC, and feel that it will be better for them than switching > vendor. Do you mean object-code compatibility? >* People who simply need, for political or psychological reasons, to > be using the large, expensive, and very "corporate" computer. They > would have bought System/Z in the nineties, rather than large Suns. Give to IBM what's IBM's and give to Intel what is Intel. >The last two groups don't tend to believe that there is a real risk of >Itanium disappearing, largely because they believe the statements to >that effect from HP and Intel executives. The problem is that (a) Intel >is now in complete control, because HP sold their IA-64 design teams and >intellectual property to Intel, so HP's controls are purely contractual >(b) Intel will, in the end, do whatever it takes to survive, and some >point, they'll get fed up of loosing money on Itanium. They should be losing money on other products? >Going back on things executives have said is comparatively easy for >corporations; they just have to fire the executives. Shareholder >lawsuits might be a different matter, but they're a question of >management of perceptions and expectations. Intel claimed when they >introduced EM64T that Itanium would be back, and that EM64T was just a >short-term diversion while development happened. Nearly three years >later, we have Core 2 Duo, and the Itanium development programme seems >to be being stretched out, with nothing radical seeming to be in the >offing. Gravity isn't letting information outside of Intel? >"Any sufficiently advanced technology is indistinguishable from a >well-rigged demo" Test harder; ask harder questions. -- |