Please create an account to participate in the Slashdot moderation system


Forgot your password?

New Itanium More Powerful, Power Efficient 51

Heir Of The Mess writes "Intel have a press release out about their new Itanium 2 Processor. The new processor doubles the performance, and improves performance per watt by 2.5 times compared to the existing single-core versions. The flagship model triples the cache and can execute 4 threads/instructions per processor enhanced by Hyper-Threading. Transistor count is a whopping 1.7 billion. Triples the previous SPEC_int_rate_base_2000 record. Retails for US$3692 for the top of the range.
So yes the Itanium crew are still pushing forward. I wonder if this could help save SGI?"
This discussion has been archived. No new comments can be posted.

New Itanium More Powerful, Power Efficient

Comments Filter:
  • by powerlord ( 28156 ) on Thursday July 20, 2006 @09:08AM (#15749157) Journal
    Considering how little buzz this article seems to be generating (even on Slashdot), it seems like this might be a case of too little, too late.

    The world has moved on from iTanic to x86-64.
  • Intel's profits fell [], according to their earnings report last night.

    Not good for an INTC shareholder.
  • Limited Scope (Score:5, Insightful)

    by tomstdenis ( 446163 ) <tomstdenis@gm a i> on Thursday July 20, 2006 @09:12AM (#15749184) Homepage
    The problem with the IA-64 platform is there are so few applications for which it is well suited. And even for problems it IS well suited it's a matter of figuring out how to extract the desired performance.

    That said, a well tuned IA-64 application can smoke the best offerings from x86 world [on both sides of the fence]. But a $3700 USD price tag may push people away. Specially since processors like the Opteron 285 are nearly half the price and way more flexible. :-)

    • Price (Score:4, Interesting)

      by jefu ( 53450 ) on Thursday July 20, 2006 @10:16AM (#15749653) Homepage Journal
      It seems quite likely that if Intel were to push the Itanium more and get it used in more systems that the demand would rise and the price would therefore drop. Of course, this might require a lot of demand, but if the performance is good enough for the price, the demand will be there.

      It might be worthwhile for Intel to find a way to drop the price enough to put these things into more places. Even give them away to visible web installations (like slashdot, fer'nstance). Get a bit more market penetration, convince some vocal people that its a good buy and it will start to take off. (I'd be glad to take an Itanium system for free for web service - even though my primary web presence is anything but big. Even better a couple of them to let my students use for compute bound projects.)

      • Re:Price (Score:3, Interesting)

        by tomstdenis ( 446163 )
        Even with a sufficient market, right now that chip is going to be expensive. It's huge and the yield can't be that high.

        But that's kinda the problem. The cost and the fact that's its a very niche processor will never make it as common place as Xeon or Opteron processors. Specially when both Intel and AMD are pushing SSE you can get a lot of the parallel SIMD like benefits of IA64 on x86.

      • Re:Price (Score:3, Interesting)

        by boner ( 27505 )
        The Itanium has horrible price/performance for web-serving... the ad-hoc nature of web-serving does not do the architecture justice. The new dual-core x64 Conroe offering would be a much better choice.

        Has anybody noticed that the Conroe effectively kills Itanium for most workloads?
        • Re:Price (Score:2, Interesting)

          by nkpatel ( 136330 )
          While x64 may be better-suited for web serving, part of the reason x64 seems to be better than ia64 on most other workloads is that you may not have the OS and software optimized for the ia64 architecture - you're still writing and compiling in a x86-centric view and I don't think compiler technologies have evolved to the point where they can fully utilize ia64 ILP.
          • Re:Price (Score:3, Informative)

            by LWATCDR ( 28044 )
            "While x64 may be better-suited for web serving, part of the reason x64 seems to be better than ia64 on most other workloads is that you may not have the OS and software optimized for the ia64 architecture"
            Or it could be the ia64 really isn't well suited to web serving.
            The ia64 has very good double precision floating point. And seems to be very good at minimal branching / very predictable branching code. In other words not web serving. The ia64 seems to have the same problem that they had with the i860 in f
            • They're unrelated. It's just another VLIW. Same as every other VLIW.
              In the same way that all RISCs are the same.

              • I know the chips are unrelated but the issues they have seem to be identical.
                You may have a point in that VLIW just isn't practical for general computing tasks.
          • As another poster mentioned the IA64 is not a general purpose processor with decent pipeline control, register renaming, branch prediction, etc.

            This means your typical random access applications [e.g. desktop] do not behave as well.

            IA64 is really meant for the case where you can handtune that 5% of the code that takes 95% of the time in assembler. Who cares if your GUI is inefficient if you're pulling 4x the FLOPS as an Opteron in your HPC application? However, for a web server where you are doing less
          • Well that was the great white hope - compiler technology evolving. It hasn't happened, and I doubt it will in the near future, at least in it's current form. The problem is that most code that we run today is inherently non parallelizable. do 5 ops on dependent data, branch, repeat. There's nothing you can parallelize there. It'll take a major paradigm shift away from serial code for us to take advantage of ia64.

            on the flip size, some media app, that has a small kernel of code that loops, is perfect for IA6
      • "demand would rise and the price would therefore drop"

        ya know that when demand rises, price rises as well. Now if they supplied more, or the public demand less then price would fall.
    • The itanium is the ultimate batch processing box. Task switches are expensive with that great fat register set, so you just want to let it work on one thing at a time.

      Superb for numerics, as you can hide all the latencies by using enough parallelism.

      One where the compiler will save you a lot of hassle, but probably not be totally optimal as there are too many possibilities for it to investigate.

      I'd certainly like to play with one. My bet is that a well-tuned FFT could be blisteringly fast.

      • For the price you pay for IA64 boxes you're going to be learing IA64 assembler fairly quick.

        From what I heard handtuned bignum code flies on the older IA64 [doing RSA-1024 private key ops in ~500K cycles].

        But when you pay $3600 for a processor you want to get the maximum benefit of it.

    • What could an x86 processor do with a 1.7 billion transistor budget? With a vast cache and/or multiple cores and/or pipelines I would think you'd get into the same performance range or maybe even better, without a (relatively) weird architecture.
      • The problem is most architectures are fairly locked down. For instance, putting another MUL engine in would be nice except MUL only reads from two registers and only outputs to two registers. A fourth pipe would raise retire latency, etc, etc, etc. In the AMD world at least most common ALU opcodes are already directpath. Larger caches could help but they raise the latency, so would raising the set-associtivity. More TLB entries are always welcomed, etc...

        As an employee of one of the chip companies I co
  • by drachenstern ( 160456 ) <> on Thursday July 20, 2006 @09:13AM (#15749193) Journal
    I can tell you that the processors are not even in production at the main server manufacturer backing the Itanium family (ie, not Intel, the other one*).

    We have seen a few proto style units roll through, but they have all had serious problems and are not running at full speed. The engineering group either cannot or will not give us a reason why these units are running crippled, but we believe it to be a chipset issue. Hopefully we will see the servers rolling through our manufacturing process within the next 60-90 days, but no management timelines have been released.

    Here's to all those of us who want that raw power and are looking to pay for it!!!

    *Won't disclose the name since I don't know if this info violates my NDA, but screw it, the public should know this stuff. Information does deserve to be known.

    • Sigh. I guess it's our best hope since your management has killed both Alpha and PA-risc for this albatross.
      Maybe Cell...
      Lots and lots of Cells...
    • Would that server making be HP by any chance? I mean the did help develop Itanium 2 []. Just about every workstation or server I've seen with an Itanium is an now, if it is HP you wouldn't be breaking the NDA by revealing common knowledge.
    • I gotta call BS on this one. I work on the development team for many of these servers. (No, I'm not allowed to tell you when they'll ship, so don't ask). There are NO chipset issues I am aware of, and the units are not crippled. I'd be interested in finding out why you think this is the case. Case in point: ahref= detail.asp?id=106071802rel=url2html-4549 []http://www 6071802>
    • by Anonymous Coward
      They are having major problems with flaky frequency shifting. The processor uses an on-silicon ammeter to control its internal clock frequency. So if you are running code that would cause its power usage to go above the rated wattage it would secretly slow the frequency to prevent that from happening. This analog method is a crapy method of power control. It requires almost no RTL and architechture changes, but tends to be falky, limits performance in unknown ways, and causes poor silicon yield. I have a fe
  • They double the performance and lower energy requirements, improving performance per watt by 2.5 times compared to existing, single-core versions.

    So does this mean that performance is 250% of the original or 350%?
  • Near, far, wherever you are
    I believe that EPIC does go on
    Once more, you opened the door
    And you're here in my heart,
    And my heart will go on and on.

    To the tune of 20 billion down the drain until it smashed into a simple AMD hack of the x86 instruction set. Yeah... It goes on... Nothing outlasts the Energiser...

  • Maybe Intel needs to sell off their Itanium division like they did with the ARM division. Concentrate on the area the are best in. Hopefully they learned some lessons from Itanium / IA64; like they did with P4 / Netburst.
    • Re:Sell off Itanium (Score:2, Interesting)

      by GnuPooh ( 696143 )
      Intel didn't sell off an ARM division. ARM has always been a separate company. They BOUGHT the StrongARM part of Digital/Compaq years ago and renamed it XScale and still have it.

      Secondly, the current Pentium-M/Core processors are decended from Pentium-III and Pentium Pro, so the thing they learned from P4 was that it was a dead-end architecture. I'm sure there are some elements here and there they "backported" to Pentium-III to make the current Dual Core 2, but you statement is not really accurate there
  • Didn't the old Itanium have something like 9MB of L2? Wow.
    • What amazes me is the 1.7 _billion_ transistor count! Thats over 1000 80486 cpus worth of transistor, or over 500 pentiums. The power ( flops, mips ) of the Itanic is pretty poor by comparison. Hell, even a single 100 MHz 486 DX4 is still decent enough in this day and age to do useful work.
      • What amazes me is the 1.7 _billion_ transistor count!

        I'm sure the majority of that is the cache. I don't have figures, but if you want an example, check out the transister count of a large DRAM chip. Sounds impressive, 'til you realize that very little of it is logic.
        • Re:Triple the cache (Score:3, Informative)

          by tomstdenis ( 446163 )
          24MiB of 6T SRAM is 1,207,959,552 transistors just for the cells. Not including tags, address decoders, etc...

          By comparison, an Opteron uses 113,246,208 transistors for the 2048+256KB of cache [assuming they use a 6T structure which I don't know for a fact since I'm not privy to the details and technically I couldn't say even if I were, so don't assume what I said is verbatim fact, yada, I hate disclaimers] and the 4MB Duo (total of 4096+128KB of cache) needs 207,618,048 transistors for its cache.

  • Bumber for Intel (Score:3, Interesting)

    by sgt scrub ( 869860 ) <saintium@yaho[ ]om ['o.c' in gap]> on Thursday July 20, 2006 @10:03AM (#15749558)
    I buy machines for our 100M product: 4 way opteron 870's with 8G Ram (2/proc) 2x73G Cheetas. Retail the box would cost about $7000. If one processor cost me $3500 life would suck. I can't see this processor doing the work of 4 duel core opterons.
    • Oh I hate to be a grammar nazi, but...

      If you were AMD's spokesman I would be cheering for Intel by now.
      You can't spell bummer in the subject; Cheetah, the product name; dual for fecks sake; and you capitalize ram, a common word, instead of Opteron, a trademarked word.

      If this is what it is to be an AMD fanboy, count me out. I'd rather be literate.
  • The same thing has happened to Microsoft with Windows Vista that has happened to the Itanium project:

    Whenever you set out to reengineer a foundation, and you throw money at it in the form of more engineers, you are asking for long delays if not outright failure.

    • Yes, but the difference is that MS holds a court verified monopoly (on desktop OS installations).

      Once Vista ships from MS (ready or not, here it comes), it WILL ship on consumer boxes (want it or not, here it comes).

      So that within a year or so after launch, most new PCs purchased will come with Vista pre-installed, especially after they stop selling XP.

      If Intel decided to stop shipping all chips but the Itanium, most PC makers would laugh, and probably abandon them to use AMD.
      • Since Itanium's team doesn't have a monopoly, as you point out, it may be they'll have sufficient motivation to make it work correctly enough to compete.

        We may yet see an Itanium that competes but I'm not holding my breath for Vista.

  • These will be very rare collectors items in a few years.
  • It's interesting the way so many people on slashdot are willing to disregard or even diss intel's Itanium strategy. Admittedly intel mistimed the itanium introduction, trying to move the processor world in a whole new direction just as AMD started producing some good chips was a mistake. Though to be fair to intel it isn't clear they could have foreseen this when they made the decisions. However, just because Itanium isn't going to run the next version of WoW or Quake doesn't mean that it is dead or a ba
    • You remember when development of this chip was announced right?

      1994. I remember the day I read the article in computerworld about the HP and Intel alliance and the new CPU, sounded interesting.

      12 years later they are shipping about 40,000 anually.

      This thing is simply not going to take off.
    • I've been under the impression that Itanium is really intended for "super-computer" type applications, as opposed to traditional desktop and web serving. For example, Itanium is used to search for oil and model areodynamics.

      Comparing Itanium to x86 is like comparing an oil tanker to a personal sailboat, or like comparing apples to oranges. x86 isn't meant for everything.

    • Admittedly intel mistimed the itanium introduction

      Mistimed? Intel had been working fruitlessly on it for 8-10 years when AMD finally realised there was an opportunity to pull a Microsoft on them (i.e. consumers like compatibility more than new tech).

      Itanium's big advantage is that it is simply a better ISA.

      Debatable. It really depends on the application. Itanium does have a nicely designed, regular ISA that is awesome for serious number-crunching, but the VLIW approach really isn't optimal for gen

FORTUNE'S FUN FACTS TO KNOW AND TELL: A giant panda bear is really a member of the racoon family.