Slashdot Mirror
Why Doesn't the Itanium Get the Respect It's Due?
happycorp wonders: "As in recent years the Itanium does well, easily beating x86 processors even at its low clockspeed (1.4Ghz). The supercomputer people are serious about benchmarking (no easily tricked microbenchmarks or reliance on closed-source
commercial apps), so the discrepancy between the performance and perception of this chip is serious.
With a single-CPU Itanium2 system at
around $2000 their price is already reasonable, and the price would come down
(and software would be ported) if the Itanium ever became a mass market chip. Having an affordable chip one step above a Xeon or Opteron in floating-point performance would not be such a bad thing for gaming enthusiasts (or 3D artists). So, the recent
article
on the
Top 500 supercomputers list brings up a question I've been meaning to ask:
Why do we see so many disparaging opinions of the Itanium processor (all those 'Itanic' jokes, etc.)?"
"It seems computing enthusiasts' sentiment is set against this processor, and its likely that it's going to be abandoned sooner or later. We'll be paying for x86 compatibility indefinitely (recall the Xeon has roughly
three times the number of transistors of the ppc970 for example; but we hardly get three times the performance).
These are a couple scores from the top 20, with the total gigaflops divided by the number of processors to obtain a per-processor speed:
rank processor ghz (gflops / #procs) speed #5 ppc970 2.2 (27910 / 4800) 5.81 #7 itanium2 1.4 (19940 / 4096) 4.86 #10 opteron 2.0 (15250 / 5000) 3.05 #20 xeon 3.06 (9819 / 2500) 3.92
Given this, consider what a 2 or 3 Ghz Itanium could do.
(fine print: I am not affiliated with the Itanium or the top500 list in any way)."
These are a couple scores from the top 20, with the total gigaflops divided by the number of processors to obtain a per-processor speed:
rank processor ghz (gflops / #procs) speed #5 ppc970 2.2 (27910 / 4800) 5.81 #7 itanium2 1.4 (19940 / 4096) 4.86 #10 opteron 2.0 (15250 / 5000) 3.05 #20 xeon 3.06 (9819 / 2500) 3.92
Given this, consider what a 2 or 3 Ghz Itanium could do.
(fine print: I am not affiliated with the Itanium or the top500 list in any way)."
I had to study the chip in one of my EE class. The technology in it is really really impressive. I love the memory architecture provisions!
One, it gets no respect because nobody uses it. Where is the kudos for the transputer? Why does nobody love the Apple ///?
Second, yes it beats the x86 into the ground. I'm not surprised. Now show me how it compares against a real CPU. We've already seen that the Itanium is competing in a different space (supercomputers), so show me how it compares with the MIPS that SGI have ditched in its favour. I wouldn't be surprised if an n GHz MIPS stuffs an n GHz Itanic into the floor.
Probably because when it mattered a single CPU Itanic was more like $12,000 and not $2,000. After fucking up all their marketing and delivering strategies no one wants one anymore.
I'm Rick James with mod points biatch!
Itanium was a huge project jointly developed with many partners, most of the significant ones have long since abandoned the effort.
It was supposed to be the future of Intel - shipping units on the order of the pentium line. A redesign from scratch of how processors "should" be designed.
It's taken far longer, cost far more, and yielded far less than promised.
That's basically it.
Also, I'd be willing to bet Intel staked a bigger part of its decision on the availablity of platform independent binaries making serious inroads, which hasn't really materialized. Platform independence of the major OSS and commerical apps is obtained through porting and source-level compatability.
but my understanding, from the rumor mill, says that the Itanium was too little, too late and was partially aborted in an effort to get it out of the lab. It was a joint HP/Intel effort that was supposed to be the "next big thing" in processors, but dragged on so long in the lab (more than 10 years) that, by the time it was released, contemporary competitors already had nearly comparable horsepower and an established mindshare.
I may be entirely wrong, but I believe the dislike for the Itanium stems from the fact that you can't compile any decently optimized code for it. Apparently, even Intel can't create a good compiler/linker and toolkit for creating machine code that makes good use of EPIC. Even though the processor itself is more efficient and faster, the same thing compiled to machine code running side by side with an Opteron or any other x86-64 chip will see the x86 win. If somebody could come up with a decent compiler/linker that provided full EPIC optimizations, they would be bangin, but they don't have it so we don't use it.
The people who work on scientific applications take performance seriously. They put a lot of effort into optimization. The itanium architecture is hard to optimize for, and the compilers just aren't there yet for the general case. So you wind up with a disparity between the performance in scientific applications and general purpose applications.
Other reasons itanium can't compete:
1) Compare the performance of itanium with xeon/opteron in running native x86 code.
2) Compare the costs of building real end user systems.
3) Compare the availability of windows xp drivers.
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
to compile for Itanium. Speaking as a compiler researcher, Itanium is great for generating research papers because there are all sorts of things that you can do from a compiler perspective. The problem is, outside a research environment, someone has to implement a lot of the ideas in an Itanium compiler to make it useful. Unfortunately, most of the stuff in the Itanium research papers isn't easy to implement and most of what gets put into commercial compilers are the easily implementable ideas.
May as well as why Linux/Mac/*BSD/etc. doesn't get the "respect it deserves." There is no real answer.
My personal thought is that price:performance was not in line with other choices available to the end consumer.
Do really dense people warp space more than others?
Itanium is broken by design. Its asm syntax etc is different from amd64 on *purpose* which was to create to incompatible 64bit system where Intel would keep AMD out of the market. Oh well, they failed badly.
Never learn by your mistakes, if you do you may never dare to try again
My systems professor told us that they chose to create a very complicated assembly language, that while may be efficient, makes programming un-nesceissarly difficult. If people don't want to program on your platform, you have a problem.
I only ever called it the Itanic because one of my professors, who works (or worked) at Intel and researched the architecture very extensively to document it also called it the Itanic. According to him, it was basically what everyone else has been saying so far.. great idea, bad execution.
It's like sex, except I'm having it!
Itanium was killed by intel's megahertz marketing. Why get an expensive 1.4Ghz itanium when you can get 2 3.0ghz xeons for less? The amd-intel 1Ghz race hit it even harder, since intel had to totally sell out itanium's higher ipc for the p3's higher frequency, and meant the p3 could be brute-forced to equal or greater performance as the new, non-mainstream itanium architecture.
In my opinion the p4 was the worst thing ever to come out of any microprocessor house in the last 20 years, as it not only comprimised microprocessor design for the horrible and blind-sighted goal of mainstream marketing, but essentially caused a large part of the current TDP crises the industry is in now, and reinforced our mentally handicapped reliance on single-threaded programming.
The humor in the itanic label has nothing to do with the chips, it has to do with intel trying to have it both ways: intel chips are the most powerful, with the only metric that matters, frequency, and ipc and design efficiency matter little, but also that "oh yeah and we have this amazing chip that is so powerful but runs at half the clock speed." It was a blatent contradiction in marketing messages.
For f*cks sake, they called their double-clocked alu "NetBurst"... seriously, why not add an onboard memory controller and claim it's "SuperBandwithMaker", which uses it's amazing technology to increase the speed of your dial-up connection...
Yes, if you market to customers by treating them as idiots, expect them to choose the stupid product, and ignore you when you claim to offer another product that "no really this is a good chip, not like that other one which we said was the fastest", which is actually better for you in the long run, because you can set a new foundation for improvement.
When amd came out with the opteron at 64-bit, and with surprisingly competitive performance while still running legacy apps at faster speeds, how do you compete with that?
Here's hoping they do manage to resurrect the alpha lines, Ibm even went a little over to the marketing darkside with the g5, trading frequency scaling for TDP, but they usually manage to rebalance the two after a few years of revisions.
The first rule of USENET is you do not talk about USENET.
I worked at a startup that was building a database ~70 gigs in size. It took 2 months to build said database. Lots and lots of very small lookups and inserts.
Memory was our bottleneck. More ram equals more speed. So we spent BIG bucks and bought a quad Itanium with 12 or 16 gigs of memory (I forget exactly how much it had).
The Itanium was slower than a dual X86 with 2 gigs of memory! And not just a little slower. We spent weeks trying to get the database optimized.
Why does no one respect the Itaniums? Intel made a slow chip. Then they released the sequel. I've already paid my dues on that line once. I'm not playing this round.
Agile Artisans
Java. Mainly.
Java is handy for many, many things. One thing it's not going to help is a large scale migration from one platform to the next. Years ago I am sure Intel thought it was going to be bigger, but it's not. Meaning more trouble migrating to new hardware platforms!
Intel was so late in delivery that all the high performance workstation people abandoned the Itanic.
Itaniums also have up to 9 Megabytes of cache. And they come with super heavy-duty FPU units. It's unfortunate that we can't directly compare the core architecture of the Itanium vs. the core architecure of the x86 without influence of the extra cache and FPU resources that the Itanium target market supports.
If somebody would build an x86 with that much cache, and increase the proportion of transistors in the x86 FPUs to match the Itanium's target market, then we could compare the relative benefits of x86 core vs. EPIC core. Until then, when you compare the two processors, you're most likely looking at differences in the logic units outside the core, which are determined by target market price/performance tradeoffs and which are largely independent of the core architecture.
I spent a while working on building cluster of IA64 machines in the HPC space. I don't anymore, and I can't say as I miss them.
First, as noted elsewhere, at the time when opinions were being formed, the cost of a decent dual processor machine with a useful amount of memory was about $30K (a typical HP cluster machine) while the comparable x86 box was about $3000. Today those prices are lower, but the x86 is still cheaper by several times.
Second, the HPC folks were perfectly happy running Linux, while no one else seemed to be. And Linux works on the machines. But in general purpose computing the advantages of the IA64 aren't seen.
Especially now when we have x86 with 64 bit memory addressing.
Next, the hardware (when you're willing to pay for it) was lousy. One box from HP had 15 fans for the 2 processors, in a 6 or 7U box. I think it would take off if you aligned it properly before turning it on. I can't imagine a rack full of these, much less a whole machine room (we had 2, and wouldn't allow them both on at the same time, but we had these in office space).
The EFI BIOS is truly an amazing piece of work. It's hard to work with on the console if you're a PC user expecting normal menu traversal with arrow keys (left and right arrow keys worked, but to move up or down you needed to use 'u' or 'd'). But we were building clusters, and expected to do things with the machine via serial interfaces. But the UI was even worse over a serial line. What happened to nice, easy to program command line firmware? (Note to HP Fort Collins: you're supposed to take the BEST of both worlds when you try to merge paradigms).
Yes, there was also a network interface which would take the place of the serial connection. But this wouldn't use DHCP, but rather manual network configuration, which required this funky hydra cable to connect to some other machine in order to set up properly. And each person connecting to this system would get write access to the console. There are decent ways to share a console across multiple logged in users (c.f. http://conserver.com/ for one example).
For a set of machines designed for cluster computing, they were about as unfriendly as you can get for us infrastructure people.
(Note, these hardware obvservations are only based on about 6 classes of HP machines that I've put my fingers onto; I have no knowledge of any non-HP IA64 machines).
Once up and running, they're not bad systems. But they're not so much better than x86 systems to win over most people. There's always a class of user who is willing to pay for the highest performance. But the larger part of the market pays attention to flops/dollar, and building a thousand node x86 system is still cheaper than 500 nodes of IA64.
(Disclaimer: I haven't had to price out IA64 systems for a year and change now; I don't believe my claims are invalid, but if someone can show me I'm wrong, I'd love to know it).
I have no interest in working on IA64 systems anymore, unless you're paying me to do it.
Oh, and it would help if Intel could make decent motherboards for its own chip.
It could be my information is out of date, but I don't believe Intel has made their own motherboards for years. It's just not cost-effective. Rather, they give specifications for boards and 3rd-party manufaturers then fabtricate their own varieties. Or at least that was the way it was when I co-oped with them. *wry grin* I always wondered how much they saved after the costs of bad publicity due to bad implementations. Amazingly often, a company would ignore corrections to their motherboard diagrams and of course Intel would get blamed when the motherboards would malfunction...
This sig has absolutely no significance and serves only to take up screen space and waste the time of the reader.
Well, there are many reasons the Itanic failed. It was a great architecture, a neat idea. Shift all of the intelligence in the chip up to the compiler, execute in-order, optimised code, get rid of deep bypassing, etc. Generally, get rid of the extra 50% of the chip that's dedicated to turning an instruction stream into a series of vectors.
Note, it *was* a neat architecture.
Then, everybody got involved. Imagine a roomfull of architecture, compiler, and systems PhD's, each with their own pet idea. And this chip had them ALL in it. Anybody remember the i432? In a way, this was the i433.
BUT. This meant a complete break with the current codebase, and in the final analysis intel didn't have the guts for it. Especially once their hopes for compilers weren't being borne out (once, Intel was a HUGE player in the market for compilers PhD's). So the guys at Intel decided to add x86 hardware compatilbility to this. Then, since their compiler plans weren't working out, they added out-of-order execution.
Now, all of these things had crazy interactions. Suddenly, who knew what it was doing? Then the power... all those units, executing all those dead instructions - it ran HOT. Then the fact that x86 compat and o-o-o were a gigantic boat anchor in terms of chip real estate, driving the cost through the roof pretty much sealed its fate. It became a "server processor". And if you get 7 or 8 P4's for the price of one Itanium... well, your cluster is better served with those 7 or 8 P4's.
Pride goeth.
I think the big problem is that it cannot run x86 software very quickly.
Yeah, that is why semi trailers don't get respect like Dodge Neons. They use diesel fuel instead of unleaded!
My point is that if your buying a 64bit system that is fast in order to run your old 32bit programs slowly. Wrong tool for the job.
I've got 65 Itanium processors downstairs. They are fast and reliable for high memory bandwidth floating point calculations, which is what we use them for. They may be a disappointment with running IE or Outlook, but for crunching numbers they are great. I have yet to of tried an Opteron but will in the next couple of weeks. From what I understand those too have become great at high memory bandwidth number crunching, but I'll wait for the numbers vs marketing speak. Now, Itaniums do suck in the power consumption and heat dissipation department.
Itaniums get such a bad rep here on Slashdot because its cool to do so. Itaniums are made by the "big guy", Intel. If they were made by AMD they would not get the same rap as they do.
The other big thing against the Itaniums is market need. A generic x86 that you can throw in the trash and replace for about $1k if there are any problems are sufficient for 99% of the servers out there. If not even preferred. Now, what other market would want a fast 64bit architecture with high memory bandwidth -- databases. Sun and Oracle fill this void. Well except for the fast and high memory bandwidth part, but Oracle+Sun is a proven combination with years of experience. Solaris does not run on Itaniums. Linux does (flawlessly), but even Oracle+Linux is not that widely adopted. I have no clue about Windows state on an Itanium. I see no real use to run Windows on an Itanium, but someone else might, but I doubt its very common.
Although Intel has some more to go with the low-voltage Itaniums because they are capped at 1.3GHz, but they are working on that. Also, Intel has dropped the price of these guys considerably. This too was an issue with Itaniums, but they have dropped by about 1/2 the price over the years.
IMHO, Intel should continue on the power management issues and price and market these chips more for number crunching. Their performance on the top500 site is impressive, but even if all of the top 500 computers used 4,000 Itanium processors each, that would only be 2,000,000 processors total, and a super computer that size is not purchased very frequently.
Most amusing to me was that the early versions had the chip serial numbers on the area covered with the heatsink. Removing the heatsink voided your warranty. You needed that serial number to get warranty work done on the processor.
This sig has absolutely no significance and serves only to take up screen space and waste the time of the reader.
Well, Without drivers it ain't worth much, is it? I'm not sure I could talk anyone into spending 2 grand & then slipping in that half the hardware they have invested in will likely never see support... and as far as gaming goes, without the drivers for the game pads\joy sticks\what ever it is they like to use, there isn't much gaming at all. Besides, you might have gotten it to install, but that is a far cry from getting it to perform acceptably with the 5 to10 grand they have invested in software. The performance didn't, and still doesn't, even come close to what they already have. Not much of an upgrade when the software runs slower... Consumer\small office VS large corporate back office are two very different worlds. Flops don't mean much to 95% of the market - if what they already have & what they want down the road don't perform at least as well on the new platform as the one they already have, your never gonna get them to buy it, even if you market the hell out of it. You all need to go back & read the why's of the fiasco - the info is all over the web.
That question answers itself: You think differently from most people. Highly specialized, hand optimized massively parallel predictable crunching seems to matter to you. It doesn't to most people. You're in a minority. Get used to it.
BTW, i860 and Alpha suffered from basically the same problem.
Itanic has its reputation because its a more fundamentally broken design than the i432 was...
PA-RISC, MIPS and SPARC had nothing going for them - no competitive advantage, certainly not over AMD64. they were each tied to old-fashioned name-brand-Unix machines, which are now thankfully extinct.
PA-RISC, MIPS, and SPARC were killed off by the mere threat of Itanium about 5 years before AMD64 first shipped. MIPS scrapped Beast and Capitan back in 1998. AMD64 was barely a glint in its designers' eyes at that point.
We'll never know if they would have been competitive. However, the one company that didn't shitcan their 64 bit RISC development, IBM, now holds a dominant position in high performance computing. Someone clearly thinks they have a competitive advantage or they wouldn't be selling supercomputers.
Martin
A couple of points that seem to have been missed when looking at why the itanium less widespread:
- each CPU is quite large, having a square surface area for the unit about 2" x 5" and it's about 2" high
- That area includes a voltage regulater and the passive cooling fans
- It doesn't include any of the necessary active cooling
If you add these physical factors to the points already made about heat, power and EFI bios, it's obvious to say that Itanium won't run in your mini-ATX destop or laptop. This isn't a slam on the design, as it was never designed to run in those form factors, but it's hard to see how any cpu today is going to have a wide use if it isn't available for dual use for destop and servers. Once you eliminate the desktop market, (and I'm going to lump the workstation market in with the servers) the number of places you can sell these processors drops considerably.Once you start adding in the lack of Windows support for itanium, the strides that the 86_64 architechture has made in capability, and the low numbers of current adopters, it's not looking like Itanium will ever gain widespread acceptance.
The Internet has no garbage collection
Why this chip is not for me are two reasons:
1: I'm not buying one before the software is ported to it -- and at a comparable price to its PC equivalent!
2: It may be a step above an Opteron for floating point, but is it still that step about a dual processor Opteron that I can buy today for less money than a mono-processor Itantium?
As for the "Itanic" jokes (all of which are way off-base, since heat output of any H.M.S. Itanic would melt any iceberg long before it could do any damage), blame The Register. I saw them use the term long before anyone else.
"It's the height of ridiculousness to say for those 9 lines you get hundreds of millions."
First off, remember that the current Itanium is the Itanium 2. The original Itanium was a failure in almost every way.
A reason why the DEC Alpha and other processors died wasn't because of a bad design, it was because the software support wasn't there. Think about it, if you could run all your Windows apps under Linux, more people would run it. The difference in what software is available is a big issue. Sure you can recompile many things to run on a given piece of hardware, but for endusers and even when it comes to servers, if a company needs to spend a lot of time making a program run well on a new architecture, and it doesn't sell well, it's a waste of resources to do it.
So, look at the current market. The Itanium 2 is horrible when it comes to performance under Windows 2003 or Windows XP. If you have native support, then the performance comes up to speed a bit better, but compare that to the performance of an Athlon 64/Opteron and the P4/Xeon chips.
Then you have the issue of how fast does it do in the real world. Theoretical performance in benchmarks is one thing, but how well does it work in a real-world environment? Can you say that the Xeon can hold it's own against an Opteron with a multi-processor system? In some applications it can, but as time goes on, AMD has tweaked the Opteron and Athlon 64 design so even in areas where Intel has dominated in the past, AMD is getting much closer in terms of performance.
Why go to a new architecture, which means all new software when you can stick to a tested architecture that runs all your current applications faster?
The Itanium 1 is like the Pentium Pro. In theory it was better because it dropped the legacy support that plagues the x86 world, but since it failed to be better at running the current/older software for the x86 world, it didn't do well. Eventually the Pentium Pro design was used in the Pentium 2, and as a result it did well, but the original implementation was horrible.
The Itanium 2 design may eventually make it's way into the current x86 world(compatability and such), but very few would be willing to switch to it.
The Opteron/Athlon 64 design is up to 2.8GHz, and is expected to eventually make it up to the 3.8GHz mark which is where the highest end Pentium 4 processor is. All things considered, I'd stick with the Opteron for multi-processor systems, or Athlon 64 FX 57 for single-processor. You get the best of all worlds without the drawbacks.
in real supercomputing you do not want your processor to 'auto-schedule' or rearrange your code.
in the end, real special code is still hand oprimized, since no compiler nor any built-in rescheduling algorithms can actually know what I really want to achieve.
Maybe I just want to accept the half ready value because I don't care for part of it.
Maybe I want to put one instruction way ahead to prime a set of registers for what is coming.
A processor which is always auto-scheduling can achieve only performance within the foresight that the rescheduling design put into it. But not for my very special algorithm for just this one dumb equation I want to solve.
It is not a 'fail' criterion for a processor to strictly adhere to what I tell it, and thus provide an exactly reproduceable solution path each time.
The result of this automatic rescheduling is that execution times in the end become non deterministic. In some cases you just want to avoid that.
The scope is for sure not gaming, but hand optimized supercomputer grade code: check the compiler's result over and manually squeeze clock cycles out by doing things that seem to put cycles in, but reward in the end because one 'senseless' instruction may just have served to prime a cache o, register file, or vector set with new content for the next run, just in time when the pipe runs empty.
Curious why the DecAlpha does not appear in these posts....
1. Power consumption,
2. Power consumption, and of course,
3. Power consumption.
Replying anonymously because I already modded in this story. Glad you brought up points 1 and 3, because I work for a company that builds custom Linux clusters for Los Alamos National Laboratories and other government agenCIes who shAll remain nameless. They use Opterons exclusively, for reasons 1 and 3. When you figure out how much it takes to power and cool 100+ node clusters and how many more compute nodes you could buy for that money, the reasons are obvious. We had our Intel rep come to us a few months back, asking what they could do to get us back to selling more Intel chips like we used to, and we said, "Cut power consumption." Number one concern, not FLOPS or anything else. Cut power consumption. Cheaper would be nice, too.
Don't forget that an IBM employee called Itanium a science project when IBM dropped there line of Itaniums.
The problem is and will remain that you don't get enough performance for what you pay for from Itanium.
When I can build a Athlon64 X2 cluster for under $500,000 that can place somewhere around 150 on the top 500 list. It's hard for itanium to compete pricewise.
Furthermore, having just tried to order a pair of Itanium 2 servers for a SQL cluster:
These supply issues quickly gathered the chip a new moniker - Unobtanium.
They exist, they fail miserably optimizing most C and C++ code or at least they aren't good enough at it to make up for Itanic's handicapped clock speed.
You just need to read the other posts here about how hard it is to develop compilers that can find 4 way parallel instructions to cram in to the VLIW at compile time. You find a lot more opportunities at runtime using dynamic scheduling at the price of complexity in the CPU.
Maybe someday the compilers will be really good and Pentirum/AMD CPU clocks will hit the wall and Itanic will reign supreme. Intel is one of the few companies with pockets deep enough to keep it alive and keep pouring the billions in to both the CPU and the compiler, until it starts outshining x86_64 on anything other than vectorizable Fortran. Wouldn't necessarily count on that confluence of events happening in time to save it. I'd really like to see how much Intel has sunk in to Itanic versus the ROI. It must be appalling. Only a company with a near monopoly elsewhere could survive it.
Me I'll take an AMD 3400+. My whole computer cost $800 versus $2000 for just an Itanic CPU, it has 2 GB/sec memory bandwidth, runs IA32 apps really fast, is running Gentoo Linux so everything is taking advantage of all the new registers and instructions set improvements, and I have 64 bit addressing. Its sweet and sensible.
I'm not argueing that Itanic wont hold its niche in supercomputing. Aren't many people who are going to put one on a desktop or in a server.
@de_machina