Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Why don't you tell us?
SP2 delayed again. Perhaps in a few more days:
http://www.siliconinvestor.com/stocktalk/msg.gsp?msgid=20379550
upc
Or maybe it's that Opteron revenue surpassed Itanium revenue a while ago... <GGGGG>
Disappointing Intel chipset performance:
http://www.xbitlabs.com/articles/chipsets/display/p4-fastest_10.html
Conclusion
I will now try to summarize. First of all, the new chipsets from Intel, although rich in new and exciting functions intended for a performance growth in the future (DDR2 SDRAM, PCI Express x16 graphics bus), are no better yet than their predecessors. Across a number of tests, the i865PE with DDR400 or DDR533 memory outperforms any of the new chipsets.
By the way, I also couldn’t see any significant advantages of the top-end i925X over the mainstream i915P. Although these products belong to different price sectors, and the i925X features certain memory controller optimizations, they both have a similar speed working with the same memory.
...
Well, the first feeling the new chipsets from Intel provoke is that of disappointment. These chipsets cannot reach new performance peaks except when the application is actively using the graphics bus. Adding the yet-unsolved overclocking problems, the i925/i915 seem to have no performance advantages from the user’s point of view. Yes, sometimes they win a test but the overall impression is anyway bad. The rather low speed of the new Intel chipsets is somewhat compensated by such features as an improved IDE controller, high-quality sound, numerous USB ports and so on. They are of little consolation, though, especially considering that the i925/i915 family will be soon replaced with new and, hopefully, faster chipsets with support of the 1066MHz FSB, and then by the Glenwood/Lakeport family.
HP sold the Itanium systems to the Weather Channel -- not Intel. So explain how you think that "Intel marketing" was involved
You didn't know that Intel helps partners win deals for Itanium, and that they use price as a bargaining instrument?
That's up there with Semi not realizing Intel has an enormous server whitebox program.
I think Intel HR needs to better educate the employees, or something.
upc
I hope it wasn't the "Prius", which, it turns out, doesn't get nearly the mileage the EPA reports, as the EPA mileage is, get this, computed based on an old *emissions* model instead of actually measuring how much gas is used! Still, probably better mileage than most cars.
http://www.idealog.us/2004/05/why_your_toyota.html
upc
ASP is it now? Wasn't it once IPF *revenues* you compared to Opteron? What happened? <GGG>
upc
You know, with the Sempron 3100+, it appears we can safely say that for gaming, the P4 has been "Sempronized". :)
upc
Isn't it obvious? The weather channel was a fairly high profile win for AMD/Opteron early on. Intel PR/marketing has the resources to focus on such deals, shovel $$$ their way, and grab back an account. They probably take a loss on those systems, but hey, they have to get folks to accept IPF systems *somehow*.
upc
Make that, rather, "On-chip memory controllers (AMD) win out over large caches (Intel)."
upc
How much Intel marketing money did it take? Were those systems effectively free to the Weather Channel?
upc
Lindenhurst problems. No systems for 30 days:
http://www.theinquirer.net/?article=17641
One large OEM went as far as saying that its launch plans were almost scrapped 24 hours prior to Intel’s announcement.
[...]
IBM, HP, as well as Dell stated that systems would not be available to ship for the next 30 days. One of Intel’s closest partners stated that shipments would not commence for at least six to eight weeks. If this timeframe is accurate as described by the OEMs, it would seem that they are waiting for stepping C4 of the Lindenhaste chipset to be available.
upc
Both Anandtech and HardOCP are recommending AMD CPUs for Doom3. Good news. Also notice that Sempron 3100+ slaughters a lot of Intel CPUs beyond Celeron D's. Like Prescott and Northwoods around the 3GHz mark. 'mas' should be pleased.
upc
dreams meet cold hard asphalt. Why are you talking about IPF again?
upc
Think a little harder. You'll get it.
Okay, maybe you won't. Here you go:
Both 32-bit and EM64T tests should have been run with 4GB of memory in the box. That would be an apples-to-apples evaluation of any benefit from EM64T mode.
And finally, raw scores should have been posted for both tests. That prevents cheating through an artificially bad performance in 32bit.
upc
I don't think it'll ever be a significant marketing point
Quite the reverse. It will be a temporary significant marketing point. Eventually, Intel will have NX on the desktop, and in the mobile space. But AMD has Q3 to itself for server/desktop, and at least H2 in mobile.
upc
How insulting! Don't suggest that I would ever act like you. <GGGG>
upc
Nice try Alan, but the memory "change" was a memory INCREASE for EM64T mode. It tells us the true results are even WORSE than the meager gains demonstrated. <ggg>
And where are the raw scores? Why didn't Intel list them?
upc
Unfortunately, those results are not apples to apples. In addition to running in EM64T mode, they doubled the memory footprint.
And no, I didn't miss them. I was correcting your claim of a 67% improvement in a "benchmark". <ggg>
upc
Semprons will not hit the channel until Aug 10. (eom)
Those are very unimpressive improvements, and the methodology has problems:
1. No raw scores given in the notes only ratios, so 32bit performance could be artificially low. No way to compare with the competition. What are they afraid of?
2. Unclear if 1 or 2 processors used, notes only say "processors". I think 2 is more likely...
3. DIFFERENT amounts of memory in the machine!!! 4GB in 32bit test, 8GB in 64bit test. This alone could be responsible for much of the improvement.
upc
There's a synthetic matrix multiple test that receives a 67% improvement
LOL! If only. What is 67% better? The *size of the problem*! LOL!!!
Compute Larger Problems
Matrix Size (Ku x Ku) - Higher is better
There's also a synthetic "amount of memory in my machine" test that's 100% better! <GGGG>
Oooh, the "8" is much larger than the "4".
What a joke! EM64T is a dog.
upc
There's a synthetic matrix multiple test that receives a 67% improvement
LOL! If only. What is 67% better? The *size of the problem*! LOL!!!
Compute Larger Problems
Matrix Size (Ku x Ku) - Higher is better
There's also a synthetic "amount of memory in my machine" test that's 100% better! <GGGG>
Oooh, the "8" is much larger than the "4".
What a joke! EM64T is a dog.
upc
Okay, here's the fine print on Intel EMT64 benchmarks:
Intel® EM64T Performance
Benefits of Intel® Extended Memory 64 Technologyφ
Source: Intel Corporation, Data current as of August 2004
Intel® EM64T technology Extended Compute Capability
Linpack* performance
* 32-bit mode: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 4 GB DDR2 400, Redhat EL3 update 2 32b Linux and 32b Linpack binary
* 64-bit mode: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 8 GB DDR2 400, Redhat EL3 update 2 64b Linux and 64b Linpack binary
Intel® EM64T technology Increased Productivity
1) Maya* v6.0 performance
* Baseline measurement: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 4 GB DDR2 400, Microsoft* Windows* XP Professional, SP1, build 2600, and Maya v6.0. Average of 5 rendering workloads
* New platform measurement: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 8 GB DDR2 400, Microsoft* Windows* XP 64-bit Edition, SP1, v.1137 build 3790, and Maya v6.0. Average of 5 rendering workloads
2) Linux Kernael Build* performance
* Baseline measurement: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 4 GB DDR2 400, Redhat* EL3 update 2 32b Linux and 32b Abaqus
* New platform measurement: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 8 GB DDR2 400, Redhat* EL3 update 2 64b Linux and 32b Abaqus
3) Amber* performance
* Baseline measurement: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 4 GB DDR2 400, Redhat* EL3 update 2 32b Linux and 32b Amber
* New platform measurement: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 8 GB DDR2 400, Redhat* EL3 update 2 64b Linux and 64b Amber
Intel® EM64T technology with Hyper Threading technology
Maya* v6.0 performance
* Baseline measurements: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 4 GB DDR2 400, Microsoft* Windows* XP Professional, build 2600, SP1 32b Maya 6.0
* New platform measurements: Intel® Xeon™ processors 3.60 GHz with 800 MHz system bus in an Intel® PF400 Server Reference Board with Intel® E7520 chipset, 8 GB DDR2 400, Microsoft* Windows* XP 64-bit Edition, 32b Maya 6.0
† Look for systems with the Intel® Pentium® 4 processor with HT Technology logo which your system vendor has verified utilize Hyper-Threading Technology. Performance will vary depending on the specific hardware and software you use. See http://www.intel.com/info/hyperthreading/ for information.
Φ Intel® Extended Memory 64 Technology (Intel® EM64T) requires a computer system with a processor, chipset, BIOS, operating system, device drivers and applications enabled for Intel EM64T. Processor will not operate (including 32-bit operation) without an Intel EM64T-enabled BIOS. Performance will vary depending on your hardware and software configurations. See www.intel.com/info/em64t for more information including details on which processors support EM64T or consult with your system vendor for more information.
Performance tests and ratings are measured using specific computer systems and/or components and reflect the approximate performance of Intel® products as measured by those tests. Any difference in system hardware or software design or configuration may affect actual performance. Buyers should consult other sources of information to evaluate the performance of systems or components they are considering purchasing. For more information on performance tests and on the performance of Intel products, reference http://www.intel.com/performance/resources/benchmark_limitations.htm or call (U.S.) 1-800-628-8686 or 1-916-356-3104.
* Other names and brands may be claimed as the property of others.
I think Big Blue likes the $$$ from the Intel Marketing fund.
upc
Intel confirms problem with Lindenhurst chipsets
http://www.theinquirer.net/?article=17594
upc
LOL. Where's the comparison to Opteron on the same tasks?
And did they dare put more than 4GB of memory in those machines?
upc
We've already figured out your usage. <GGG>
I understand. It's a moot point, no doubt. <gggg>
Too bad Nocona 64-bit sucks. More info:
http://www.infoworld.com/article/04/08/02/31OPcurve_1.html
Ahead of the Curve
Nocona pulls a fast one
Intel asks customers to put their faith in a stopgap architecture
By Tom Yager August 02, 2004
As a Xeon, Nocona is a honey of a power boost: fat cache; fast clock speed; smaller, denser die; and faster front-side bus. Intel’s engineers busted hump to make Nocona Intel’s best Xeon CPU. It’s a shame that this laudable effort will be overshadowed by the last-minute inclusion of EM64T, Intel’s 64-bit extended memory architecture. The Nocona EM64T-equipped server CPU, and its desktop counterpart Prescott, are precisely the “64-bit instructions strapped to 32-bit CPUs” that Intel warned against when Pentium 4 Xeon was launched.
Were EM64T sold for what it is — x86 with a wider address bus and larger register set — I’d have no complaints. But Intel hopes to use EM64T to blunt the market’s growing interest in genuine 64-bit value servers and power desktops. These aren’t waiting around the corner. They’re here now, constructed around AMD Athlon 64 and Apple/IBM/Motorola PowerPC technology. But make no mistake: A 64-bit system requires a new system architecture (as AMD and Apple have created), not a new processor.
What characterizes a 64-bit system? Multiple independent I/O busses, on-chip memory controllers, and very highly integrated peripheral chip sets that eliminate the need for the many discrete single-purpose chips common to PC designs — along with the slower, noisier circuits between them. Sixty-four-bit systems are technically simple compared with PCs’ convoluted, legacy-bound designs. It took AMD several years of obsessive engineering to create Hammer (the former code name for the Athlon 64/Opteron line). Apple had to create a new 64-bit system, almost entirely of its own design, to build around the PowerPC 970; Apple earned the right to put the G5 brand on the architecture.
Hammer and PowerPC G5 have a vital trait in common that Nocona does not share: They were engineered as legitimate, pure 64-bit architectures. When this approach is taken with customers’ needs in mind and applied to the entire system, 32-bit compatibility naturally comes along for the ride. This is not an objective one can achieve by building on top of an architecture that is already well beyond the limits of its design. When Itanium was engineered, it was clear that x86 was heading for the wall. Future applications would demand high speed outside the CPU die as well as inside, performance characteristics that were, at the time, seen only in very expensive 64-bit systems. The incredible foresight behind that strategy was so badly botched in execution that when Xeon’s term expires, Itanium will not be in position to take over.
So here we have Nocona, Intel’s would-be placeholder. Intel said in advance that it doesn’t believe in this architecture. Neither do I. Nocona is, at best, a 32-bit Opteron emulator. Nocona will do well enough on 32-bit published benchmarks for a while to make those who aren’t readers of this column wonder what all the Opteron and G5 fuss is about. But during the next six to 12 months, while operating systems and I/O-intensive apps start flipping the 64-bit switch, users who already have true 64-bit AMD and G5 systems will get the total performance designed into these architectures. Those with EM64T systems will get what Intel designed into their systems, too. Why, they’ll ask, didn’t someone tell them nothing would happen when they flipped Nocona’s 64-bit switch? Isn’t it fun to know something that others don’t?
http://www.infoworld.com/article/04/08/02/31FEnoconadetail_1.html?HARDWARE
[...]
The Nocona feature that has grabbed the most attention, Intel’s new EM64T (Extended Memory 64 Technology), might be the least interesting. EM64T breaks the 4GB RAM barrier associated with all x86 processors (except Nocona, Prescott, and AMD’s Athlon 64, Athlon 64 FX, and Opteron chips). Basically, Intel created a hack for using chunks of memory above the 4GB mark that will effectively reduce the system’s contiguous address space. Yes, developers will be happy to see a big address space for their new 64-bit apps, but Intel’s system architecture will likely hamper, not improve, performance as more RAM is added to the system.
In the end, Nocona is a Xeon killer not an Opteron killer. The fine points of chip architecture aside, differences at the system level relegate Nocona to the dual-processor bush league. For example, Intel could not match the glueless I/O system. Processors in an Opteron server talk directly to one another without going through the silicon intermediaries forced by Intel’s design. Likewise, Opteron links each CPU directly to memory, and Opteron systems’ memory bandwidth scales upward with the number of CPUs. Nocona’s bandwidth remains static.
Nocona’s design is Intel’s best yet. It’s a beautiful Xeon. But to catch Opteron, Intel would have to ditch its time-honored system architecture. Given where Itanium has (or hasn’t) gone, and the incredible pace at which AMD is advancing its technology, Nocona looks like a well-manicured rest stop alongside the very bumpy road that Intel faces.
I didn't think you were "unclear", I thought you were incorrect, and given the plentiful past discussions on this point, it was hard to imagine that you were merely ignorant of the truth.
upc
Another data point on Nocona (lack of) performance:
Re:We only have c't saying EM64T 'sucks', plus the circumstantial
By Jens Nurmann on Monday, August 2, 2004 1:27 AM EDT
The last issue of c't gave some 64bit benchmarks on Linux. They used IBMs DB2 for some midsize database benchmarking and the usual LAMP tests. I am still somewhat puzzled by the results! In a nutshell
* on DB2 benchmarks 64bit (OS and application) showed no performance gains on Nocona or Opteron. That i find most surprising - still not sure wether the DB2 demo version they got from IBM really is a 64bit app.
* LAMP showed hughe improvements when going to 64bit. Nevertheless it showed much more improvements when run on Opteron than on Nocona. Ratio was >2 : 1 under 64bit while Nocona was ahead of opteron under 32bit.
Cheers,
Jens
http://www.aceshardware.com/forum?read=115092725
Actually, you didn't mention fusing until several posts after the misleading one.
The S775 chip mentioned here for Q4 launch will be Prescott based and have a model number of 720J. It will have both XD and probably EM64T.
That chip is not the one to which I referred!
(But, that one will have XD, but probably not EM64T, unless Intel changes their public position regarding EMT64 activation timing outside the server/workstation segment. And that chip may well have slipped into Q1-- An Intel spokesman said larger cache, faster FSB parts were coming NEXT YEAR, on Friday. Now, that doesn't explicitly rule out their arrival this year, but I expect the old 130nm EE stuff on 775 is more likely for this year than a new 90nm Prescott.)
I'll say it again:
They are also preparing to launch 130nm Northwoods (not EEs) on LGA-775, presumably due to the disaster that is 90nm Prescott.
Here's a link: http://www.xbitlabs.com/news/cpu/display/20040731144157.html
"In a rather surprising manner Intel also seems to move a number of 130nm chips into its new Socket T infrastructure: the firm plans to release Pentium 4 2.80C, Pentium 4 3.00C and Pentium 4 3.20C processors with 800MHz processor system bus and 512KB of L2 cache. For higher-end market segment the company will offer Pentium 4 Extreme Edition processors at 3.20GHz and 3.46GHz with 1066MHz Quad Pumped Bus and 2MB of L2 cache along with “officially planned” Pentium 4 3.80GHz with 800MHz bus and 1MB of cache."
My opinion is that Intel is stalling EMT64 outside of the server/workstation market because:
(1) The current implementation in the Prescott core has crappy performance.
(2) The more parts in the wild supporting iAMD64, the faster the potential for IPF market expansion is crushed.
upc
I don't think they will see it at all. Ever seen a retail pc display?
http://www.siliconinvestor.com/stocktalk/msg.gsp?msgid=20363318
PS: I stopped at CompUSA. There were 3 desktop Athlon64 systems, including [3400] from HP labeled as "Best Seller"
On one of the Athlon64 computers, there was a metal attachment that advertized "Enhanced Virus Protection with XP SP2.
upc
That's not the only piece of misinformation lately. There was the small matter of the suggestion that anyone buying a 32-bit Prescott now would have it magically "enabled" into 64-bit mode at some future date. Of course, Intel has ruled this out.
Intel will eventually transition to Prescott core for Extreme Edition ...
Read more carefully. They are also preparing to launch 130nm Northwoods (not EEs) on LGA-775, presumably due to the disaster that is 90nm Prescott. Those are much higher volume parts than the EE, and they will not have NX.
Of course, the biggest NX/Intel issue of all is probably Dothan's lack of that feature until sometime next year.
upc
Overhyping? I don't think so. Just another important feature that Intel is behind on. Itanium? That's a niche product. No one cares. I'm talking about the x86 / x64 (as Microsoft calls iAMD64 now) market.
You think the average customer will see "enhanced virus protection" and not realize that is a good thing?? Take off the blinders.
upc
More evidence that Intel iAMD64 performance sucks:
http://www.aceshardware.com/forum?read=115092693
Can you summarize?
By Trinity on Sunday, August 1, 2004 5:19 PM EDT
I had posted here a link to an interview with a Senior VP of Windows development who pretty much said that Intel's implementation wasn't up to the mark compared to AMD's. I'm sorry but I can't seem to find that link. If others can then please post here. Also there have been many discussions on RWT about the Linux folks talking about issues with Intel's implementation. Search for "bounce buffers" in relation to AMD64 and EMT64. I'm not totally aware of the details but it has something to do with 64 bit DMA's in IO devices like PCI cards etc and them being able to access a 64 bit address space or something like that. If anyone else knows the details or links, kindly post them here.
Edit: I found a link about the issues the Linux folks talked about -
http://lwn.net/Articles/91870/
"On some architectures - notably AMD's x86_64 - an I/O memory management unit (IOMMU) is provided. This unit remaps addresses between the peripheral bus and main memory; it can make any region of physical memory appear to exist in an area accessible by the device. Systems equipped with an IOMMU thus have no problems allocating DMA memory - any memory will do. Unfortunately, when Intel created its variant of the x86_64 architecture, it decided to leave the IOMMU out. So devices running on "Intel inside" systems work directly with physical memory addresses, and, as a result, the more limited devices out there cannot access all of physical memory. And, as we have seen, the kernel has trouble allocating memory which meets their special needs."
Edit 2: Okay I found the other link from Winsupersite also which make it difficult to find archived articles easily -
http://www.winsupersite.com/showcase/muglia_winserver.asp
"Paul: Are you seeing any difference between AMD's [64-bit] stuff and Intel's stuff?
BM: Yes. [Smiles]
Paul: Would you care to clarify that? [Laughs]
BM: Well, AMD has done a good job ...
[Laughter]
Paul: OK, I realize these companies are both important partners...
BM: I think both have invested very heavily... and I'm sure that customers will be happy with either solution.
Paul: All righty.
[Laughter]
BM: Are there differences? Yes, there are differences.
Paul: OK, so how do these companies differentiate their 64-bit products?
BM: So there are some things that AMD's done that Intel hasn't done, and I'm sure Intel will continue to invest here, and will do a really good job. AMD led the way on this one. There's no doubt they led the way on this one."
He'll probably blame Microsoft, not Intel or AMD. (eom)
I'm sorry, did I say 'majority'?
Again, why is Intel doing this (but late) if it isn't a good thing? You can't have it both ways.
upc
Good thing NX is selectable on a per-application basis, then, isn't it?
upc