PDA

View Full Version : 8 Intel Cores Faster Than 16 AMD(Barcelona) Cores?


Pages : [1] 2

nekrosoft13
05-25-07, 09:48 PM
At the Microprocessor Forum, Intel demoed its V8 workstation running the POV-Ray benchmark. The machine is equipped with Intel's Workstation Board S5000XVN, 2 quad-core Xeons 5365, clocked at 3GHz and 16GB RAM. And the results are simply impressive: Intel scored over 4,900 pixels per seconds versus a little bit over 4,000 for AMD's 4 sockets quad-core (Barcelona) system. Again, this is an AMD 16 cores system versus Intel's 8 cores V8 machine.
If you take 16 cores at 1.8Ghz, divided by 2 and add 60% to get to 3.0Ghz, it gives POV-Ray score of 3600 for 8 cores. Meaning that at 3.0GHz, Barcelona still lose by quite some compare to an 8 cores Clowertown system.





"Why do you need 16 cores, when you can do better with 8. Our 8 core system is 30% faster than the 16 core machine AMD showed to the press yesterday. I just don't understand how they can claim to be 40% faster", said Francois Piednoel, an Intel engineer present at the show.
Well, now AMD has some explanation to do and sooner rather than later. Because if they can't figure out what happened with those POV-Ray results they showed us, that's the end of it. At the show, Intel also demoed a system with a 45nm Penryn quad-core processor (shipping by year end, about the same time than Barcelona!) that is 40% faster than the top of the line quad-core generation, the Core 2 Extreme processor QX6800. Wow!



http://www.uberpulse.com/us/2007/05/intel_responds_to_amd_barcelona_has_a_problem_you_ better_fix_it_before_you_ship.php

RussianHAXOR
05-25-07, 09:51 PM
That math just confuses me.

Redeemed
05-26-07, 04:23 AM
Chris over at AMDZone has some interesting points about these two tests. It's worth a read:

Ok, before this Povray benchmark story starts to spiral out of control I'm deciding to do the real work necessary to decode it. No, we didn't get any sort of invite to any Barcelona demonstration. Oh well. I have plenty of things cooking at once. On to what we do know.

The Inquirer has photos and a description of a Povray benchmark test pitting two 4 Socket Opteron systems, one Barcelona quad cores, and one current dual core Opterons so 8 cores versus 16 cores. We must expect that Povray 3.7 beta was used with the clock turned back on the systems because the current beta expired on May 10th. We know that the built in Povray benchmark was used. We know that pixels per second (PPS) was used, and that the rendering was shown during the benchmark which will slow down performance somewhat. We know the 4S dual core system scored about 2200PPS, and that the quad core scored just over 4000 PPS. That is what AMD said. We do not know what the default resolution rendered was. I know that my low power X2 4600+ at 2.4GHz gets just under 600PPS with 2GB. That is with running multiple Internet Explorer and Fire Fox tabs and windows. Along with Outlook, Smartlaunch, Quickbooks, and Trillian, etc, etc. 8 times that would be 16 cores, and would be 4800PPS based off a 2.4GHz number with 1GB of memory per core.

Now Intel has thrown their hat in with a 3GHz 2S quad Xeon setup with 16GB of memory and they get 4933 PPS. We don't know what resolution they ran at, and we don't know if it displayed the rendering while running or not. I'll assume it did. My 8 times what my dual core under load does is 4800, AMD's Barcelona 16 core is just over 4000, and they are at over 4900 for Xeon 2S quads. So how is Intel getting so much performance out of 8 cores, and more than AMD's 16 core? Over twice the performance is what it looks like.

Our friends at Tech Report show the quad core Intel QX6800 losing to a 4 total core FX74 setup here in Povray using the chess benchmark, and just beating the FX74 in the built in benchmark. How that translates to the current Xeon doing twice as much performance either has to be tied to the amount of memory used, or the benchmark setup itself. Clearly the current Xeon should not perform more than twice as good as the current Opteron in Povray. For more clarification we looked to AMD's Damon Muzny and we have this exclusive response.

"The objective of our demo was to show performance scaling from our current dual-core processors to our upcoming quad-core processors within the same thermal envelope and drive home the point through a real-world demonstration that customers could expect to see 2x the performance without an increase in power consumption.

At some point before our launch you can plan on us showing a demo of our parts vs. Intel's high-performance processors. For now, note that the Opterons used were HE processors.

One thing to note, the system we showed, while it was a 4P, it was running only 6GB of memory."

So we have a 10GB advantage on memory amount for the Xeon. I'm not sure if that would contribute, but I can guess it would. Also the clock speed certainly would matter, and AMD has not released what clock speed their demo was running at. Lastly we don't know what resolution, or what parameters were used in either benchmark from AMD or Intel. The data is not complete. If this hardware enthusiast was at the demo I'm sure we would have more answers. For now one thing is certain. We don't have a clear picture of how Barcelona will perform. Like wise when putting benchmarks out I would suggest a better clarification of system specs or speculation will run rampant. Speculation that will be harder to control if transparency is not present.


http://www.amdzone.com/modules.php?op=modload&name=News&file=article&sid=7794&mode=nested&order=1&thold=1

Here's the techreport link Chris mentioned:

http://techreport.com/reviews/2007q2/core2-qx6800/index.x?pg=9

In short, there is a lot of reason as to why the tests cannot be compared. They were not identical, thus meaning very little. Furthermore, if AMD were using their lowest end Barcelona chip for that demonstration- then Intel would not have much in the way of bragging rights. Too much of the picture is missing to make a complete decision.

radekhulan
05-26-07, 06:29 AM
Chris over at AMDZone has some interesting points about these two tests. It's worth a read:

..
We don't have a clear picture of how Barcelona will perform

AMD is, again, full of bull****. After huge loss in Q1/2007 they come up within 2 days with "Barcelona will be 40% faster", but do not allow anybody to see a real benchmark. After R600 fiasco and very poor performance in many games and possible Barcelona delays they come up within 2 days with "Intel is copying AMD" (source (http://www.tgdaily.com/content/view/32076/136/)) bull****, ignoring that all x86/SSE/SSE2/SSE3 architecture is taken from Intel, HyperTransport is not their invention, and on-die memory controller does not have any tangible benefits (not even for latency) in real world.

Reminds me of ATI, they did the same, when NVIDIA had fully working 2x6800 SLI on the market, they paper launched Crossfire, but did not deliver it till 2 years later, and with horrible concept (they finally copied NVIDIA internal bridge approach in 2900XT), and when NVIDIA had 7800 on the market, they "paper launched" X1800, saying how great it will perform, only to find, months later, it was a fiasco.

I would never ever believe companies doing marketing first, and products as an afterthought.

Redeemed
05-26-07, 06:41 AM
AMD is, again, full of bull****. After huge loss in Q1/2007 they come up within 2 days with "Barcelona will be 40% faster", but do not allow anybody to see a real benchmark. After R600 fiasco and very poor performance in many games and possible Barcelona delays they come up within 2 days with "Intel is copying AMD" (source (http://www.tgdaily.com/content/view/32076/136/)) bull****, ignoring that all x86/SSE/SSE2/SSE3 architecture is taken from Intel, HyperTransport is not their invention, and on-die memory controller does not have any tangible benefits (not even for latency) in real world.

Reminds me of ATI, they did the same, when NVIDIA had fully working 2x6800 SLI on the market, they paper launched Crossfire, but did not deliver it till 2 years later, and with horrible concept (they finally copied NVIDIA internal bridge approach in 2900XT), and when NVIDIA had 7800 on the market, they "paper launched" X1800, saying how great it will perform, only to find, months later, it was a fiasco.

I would never ever believe companies doing marketing first, and products as an afterthought.

Coming from the person whom has openly stated that they hate AMD and refused to ever use any of their products again, while also stating that the dual core P4 ran cooler and performed better ovar-all than the A64 x2. Yup.

buffbiff21
05-26-07, 08:50 AM
on-die memory controller does not have any tangible benefits (not even for latency) in real world.

Async upclock/downclock of RAM allowing for much more flexible overclocking?

okay.. :POKE:

it's more versatile of any of Intel's solutions even as of now.

radekhulan
05-26-07, 10:58 AM
also stating that the dual core P4 ran cooler and performed better ovar-all than the A64 x2. Yup.

No. I stated that my P4 HT 3.2 GHz performed way better for me than my Athlon 64 Venice 3200+ (yes, I owned them both), while AMD choked in ffdshow at 90% processing video realtime, P4 HT was at 20%. X2 was better (had X2 4400+), but extremely overpriced, costing more than my current Q6600, and being buggy (remember AMD dualcore drivers, too fast or too slow games without them; plus KB896256?). But that is a whole different thread...

Anyway, ATI did engage in "creative marketing" a lot (Crossfire, X1800, R600), now AMD is doing that as well. They do not have a product on the market (Barcelona), so they at least have to spread FUD, rumours, etc.

I prefer companies that can actually deliver (Intel, NVIDIA) over those "creative" ones (AMD/ATI combo) in marketing only. And yes, I will never ever buy unstable, overpriced and slow AMD product again, I can make a mistake (Venice 3200+ -> X2 4400+), but I prefer not to repeat it.

Sazar
05-26-07, 11:06 AM
No. I stated that my P4 HT 3.2 GHz performed way better for me than my Athlon 64 Venice 3200+ (yes, I owned them both), while AMD choked in ffdshow at 90%, P4 HT was at 20%. X2 was better (had X2 4400+), but extremely overpriced, costing more than my current Q6600, and being buggy (remember AMD dualcore drivers, too fast or too slow games without them; plus KB896256?). But that is a whole different thread...

Anyway, ATI did engage in "creative marketing" a lot (Crossfire, X1800, R600), now AMD is doing that as well. They do not have a product on the market (Barcelona), so they at least have to spread FUD, rumours, etc.

I prefer companies that can actually deliver (Intel, NVIDIA) over those "creative" ones (AMD/ATI combo) in marketing only. And yes, I will never ever buy unstable and overpriced and slow AMD product again, I can make a mistake (Venice 3200+ -> X2 4400+), but I prefer not to repeat it.

Radek, all the companies you listed have engaged in creative marketing over the course of about 2 years. Saying otherwise is just using selective memory my friend.

Re: The products you are listing and your experiences, I am a little surprised because my old 3400+ sct 754 proc stomped all over my 3.0c northy.

Also, I fail to understand your logic for saying the integrated memory controller is not useful. Intel will be using one in their nehalem processor and if the controller was not useful, they would not be taking this step.

Mr_LoL
05-26-07, 11:14 AM
Come on AMD show the public some benchmarks. What do they have to hide? Barcelona should blow away clovertown.

radekhulan
05-26-07, 11:20 AM
Also, I fail to understand your logic for saying the integrated memory controller is not useful. Intel will be using one in their nehalem processor and if the controller was not useful, they would not be taking this step.

I am not saying integrated memory controller is not useful. It probably is, if done properly, but in real life applications it means very little % performance gain anyway.

AMD had a very bad / buggy on-die memory controller, first batch of Athlon 64 had problems with almost any DDR modules, it did not work with 4x1GB, etc. These CPUs were simply faulty, AMD issued new revisions, but did not exchange faulty ones. Intel memory controller, on the other hand, was/is much better, and is very fast, if you measure memory read / write / latency via Sandra or Everest, you will get very similar numbers for AMD / Intel, eventhough AMD is on-die...

If Intel moves their great memory controller on-die, then we might see some benefits (5-7%?). On the other hand, we will loose simple memory upgrades by just using different board (P965 -> P35; DDR2->DDR3).

Trademark
05-26-07, 11:29 AM
Argh, redeemed, you constantly try to find some sort of excuse or explanation to why intel's benchmark sucks and why AMD doesn't, no offense, but AMD is the one being all quiet about their benchmarks, while intel openly shows what they have done. And whyt all the worries if Barcelona is not running at it's maximum potential...

sammy sung
05-26-07, 11:48 AM
Argh, redeemed, you constantly try to find some sort of excuse or explanation to why intel's benchmark sucks and why AMD doesn't, no offense, but AMD is the one being all quiet about their benchmarks, while intel openly shows what they have done. And whyt all the worries if Barcelona is not running at it's maximum potential...

Idd his glasses seems to have a slight green tint to them :p

Sazar
05-26-07, 12:14 PM
I am not saying integrated memory controller is not useful. It probably is, if done properly, but in real life applications it means very little % performance gain anyway.

AMD had a very bad / buggy on-die memory controller, first batch of Athlon 64 had problems with almost any DDR modules, it did not work with 4x1GB, etc. These CPUs were simply faulty, AMD issued new revisions, but did not exchange faulty ones. Intel memory controller, on the other hand, was/is much better, and is very fast, if you measure memory read / write / latency via Sandra or Everest, you will get very similar numbers for AMD / Intel, eventhough AMD is on-die...

If Intel moves their great memory controller on-die, then we might see some benefits (5-7%?). On the other hand, we will loose simple memory upgrades by just using different board (P965 -> P35; DDR2->DDR3).

First, you are comparing a memory controller mechanism which is several years old v/s the newer core controllers. The raw bandwidth of the processors is much improved, as are the stages. This is to be expected. I further expect penryn to continue to provide optimizations that will benefit the speed of the overall system.

Prescott and Northwood did not exhibit efficient memory management at all.

You will most likely have a new board requirement when nehalem comes forth, but I don't think that should be a painful step if the product is good quality.

Sazar
05-26-07, 12:16 PM
Argh, redeemed, you constantly try to find some sort of excuse or explanation to why intel's benchmark sucks and why AMD doesn't, no offense, but AMD is the one being all quiet about their benchmarks, while intel openly shows what they have done. And whyt all the worries if Barcelona is not running at it's maximum potential...

First off, Intel did the exact same thing before they came forth with their new design :)

Till core2's started showing their muscle-power, Intel was pretty tight-lipped.

Also, the A revisions of the barcelona procs had issues with the memory controller and bios and the B revisions seem to be having issues too. I seriously doubt AMD has a fully-specced configuration running @ 100% that they can demo.

Trademark
05-26-07, 01:12 PM
Then they shouldn't do demos untill they can show something more interesting, nobody wants to see another flawed cpu, because it it's truly that lame then it will be another Pentium 4/Athlon era, only difference is AMD being owned instead of intel.

And with the way AMD delays stuff and all the vague crap their bringing our isn't helping anyone.

As for the Core solution from Intel, that's a different story, Intel needed a comeback and AMD didn't see it comming, while Intel can see Barcelona comming, Intel is not worried because they are ahead of AMD and thisw is a great advantage, AMD will most likely steal the performance crown from the C2D for a couple of months, but what use is it when intel takes it back shortly after and AMD has nothing again?

Madpistol
05-26-07, 01:14 PM
Wow... this isn't good is it?

AMD had better not go under. I want to see some true performance benchmarks though. This stuff can be faked a little though.

However, Intel did this with the Core 2's when they were being developed. They openly compared a 2.93Ghz Core 2 duo against a 3.0 Ghz OC'd FX-60, and the Core 2 Duo walked all over it.

Looks like Intel may win again unfortunately. AMD needs to get thier act together. We need some competition so that prices will drop dramatically.

Trademark
05-26-07, 01:20 PM
AMD just has to stop, think and get something good out there, not a vague product wich is postponed every month...

jcrox
05-26-07, 01:47 PM
So we have a 10GB advantage on memory amount for the Xeon. I'm not sure if that would contribute, but I can guess it would. Also the clock speed certainly would matter, and AMD has not released what clock speed their demo was running at. Lastly we don't know what resolution, or what parameters were used in either benchmark from AMD or Intel. The data is not complete. If this hardware enthusiast was at the demo I'm sure we would have more answers. For now one thing is certain. We don't have a clear picture of how Barcelona will perform. Like wise when putting benchmarks out I would suggest a better clarification of system specs or speculation will run rampant. Speculation that will be harder to control if transparency is not present.

Before all of us "experts" make the world's final decision regarding AMD vs Intel maybe we should just wait for the chips to actually be released and make the comparison between the 2 companies on real numbers.

As for my position, when Barcelona comes out(unless it for some reason totally sucks) it will all come down to price. Barcelona doesn't necessarily have to "win" vs Intel for me to purchase one, just be competitive both performance wise and price wise....an extra 10 or 20 fps when you're already at 80+ fps means nothing to me, running any program in 2.5 seconds instead of 4 seconds is also pretty irrelevant.

radek... why even post on the topic? We're all well aware of your hate for AMD, Barcelona could be 10x faster than any Intel processor and you still wouldn't have anything good to say about it so what's the point of saying anything at all if you can't have even the slightest bit of objectiveness in your comments. Maybe you could start a "Vent your hate for AMD here" thread and get all your issues out, might make you fell a little better.

Redeemed
05-26-07, 03:28 PM
Argh, redeemed, you constantly try to find some sort of excuse or explanation to why intel's benchmark sucks and why AMD doesn't, no offense, but AMD is the one being all quiet about their benchmarks, while intel openly shows what they have done. And whyt all the worries if Barcelona is not running at it's maximum potential...
Really? Okay... so then what resolution was the Intel benchmark ran at? And the AMD benchmark? Also... are you implying that the extra 10GB of RAM had no impact on the scores the Intel setup managed? No, Chris explained the situation perfectly- I even quoted it here so ya'll didn't have to click on the link.

Also... care to clarify this:

Our friends at Tech Report show the quad core Intel QX6800 losing to a 4 total core FX74 setup here in Povray using the chess benchmark, and just beating the FX74 in the built in benchmark. How that translates to the current Xeon doing twice as much performance either has to be tied to the amount of memory used, or the benchmark setup itself. Clearly the current Xeon should not perform more than twice as good as the current Opteron in Povray. For more clarification we looked to AMD's Damon Muzny and we have this exclusive response.


From that quote from Chris Tom again. I don't think you even read that quote. And no, I'm not an AMD fanboy in the least. I'm just not so frikken' blind as to think Intel is invincible and that AMD will never catch up or surpass Intel. You guys on the other had talk as if you'd be pleased if AMD went under. What a bunch of dumbasses if that is how you feel. Like it or not we need AMD, even if they remain the underdog for a while. Atleast it's competition against Intel- which is a good thing.

God forbid somebody look at the situation with a level head, being skeptical and analyzing what they read instead of just praising intel and booing AMD. :rolleyes:

Trademark
05-26-07, 04:43 PM
I do not hate amd and I certainly don't want them to die, but they are currently killing theirselves, it's absolutely stupid to praise something like a Radeon 2900XT in comparison with a G80, it's stupid to compare a AMD cpu with an intel one when AMD is all but clear on the specs wich they showed their demo on. Tell me, why is AMD getting their asses fried? Because they got nothing, right now, untill Barcelona is released and/or/if ATI fixes it's R600 drivers then we'll know the answer, right now AMD is getting their asses fried, wether it is by intel or Nvidia, the R600 is nothing spectactular and Barcelona = vague.

As for the 10GB in memory difference, I thought Core2 based cpu's didn't really rely on memory BW etc, so what's up with that. Still doesn't change the fact that by the time Barcelona is released Intel will have another cpu ready to own some stuff

I repeat, I DO NOT HATE AMD AND I DO NOT WANT THEM TO GO DOWN, but right now they can't do **** about intel.

Viral
05-27-07, 02:11 AM
AMD is, again, full of bull****. After huge loss in Q1/2007 they come up within 2 days with "Barcelona will be 40% faster", but do not allow anybody to see a real benchmark. After R600 fiasco and very poor performance in many games and possible Barcelona delays they come up within 2 days with "Intel is copying AMD" (source (http://www.tgdaily.com/content/view/32076/136/)) bull****, ignoring that all x86/SSE/SSE2/SSE3 architecture is taken from Intel, HyperTransport is not their invention, and on-die memory controller does not have any tangible benefits (not even for latency) in real world.

I would never ever believe companies doing marketing first, and products as an afterthought.

Indeed, that doesn't sound like something Intel would ever do. They were, after all, first to market with x86-64, cool 'n' quiet/speedstep and an on-die memory controller (even though they show NO gains what so ever, of course!). Oh what about the good old days of RAMBUS, AMD were stupid to go with DDR, thank god Intel saw the light and now we're all using amazing XD-RAM today.

Seriously now, you think AMD are the ones who put marketing first and products second? You need help. AMD have traditionally never had a strong marketing force, they sink practically all of their money into R&D. That fact is not something which can be said for Intel. AMD was simply sharing some numbers. It was a closed comparison, between Barcelona and current Opterons. Why is it AMD's fault that Intel decided to throw all basis for fair comparison out the window and confuse the hell out of everyone?

As for SSE instructions. You do realise there's no possible way for AMD to sucessfully create their own instruction sets with the marketshare they have? They aren't copying Intel, they're accepting the inevitable fact that Intel has the say in industry standards like these because they have the marketshare, and thus are implementing their solutions the best they can, 100% legally and legit. This is one of the main reasons it's just so hard for AMD to ever get ahead, they really do have an impossible challenge on their hands.

Redeemed
05-27-07, 02:21 AM
Indeed, that doesn't sound like something Intel would ever do. They were, after all, first to market with x86-64, cool 'n' quiet/speedstep and an on-die memory controller (even though they show NO gains what so ever, of course!). Oh what about the good old days of RAMBUS, AMD were stupid to go with DDR, thank god Intel saw the light and now we're all using amazing XD-RAM today.

Seriously now, you think AMD are the ones who put marketing first and products second? You need help. AMD have traditionally never had a strong marketing force, they sink practically all of their money into R&D. That fact is not something which can be said for Intel. AMD was simply sharing some numbers. It was a closed comparison, between Barcelona and current Opterons. Why is it AMD's fault that Intel decided to throw all basis for fair comparison out the window and confuse the hell out of everyone?

As for SSE instructions. You do realise there's no possible way for AMD to sucessfully create their own instruction sets with the marketshare they have? They aren't copying Intel, they're accepting the inevitable fact that Intel has the say in industry standards like these because they have the marketshare, and thus are implementing their solutions the best they can, 100% legally and legit. This is one of the main reasons it's just so hard for AMD to ever get ahead, they really do have an impossible challenge on their hands.

Wow. :bugeyes:

QFT bro. Don't think I could have put that any better.

Sazar
05-27-07, 02:56 AM
Indeed, that doesn't sound like something Intel would ever do. They were, after all, first to market with x86-64

Intel was first to market with a 64 bit proc and STILL has the only TRUE 64-bit proc.

cool 'n' quiet

No question there.

and an on-die memory controller (even though they show NO gains what so ever, of course!).

Intel actually designed a memory controller way back in the day. Feel free to google it. They have experience with the design and implementation of an on-die memory controller. They just never came to market with it because at that time, there did not seem to be a need for it.

Oh what about the good old days of RAMBUS, AMD were stupid to go with DDR, thank god Intel saw the light and now we're all using amazing XD-RAM today.

Fwiw, with all the issues and bad press Rambus has had, there were some very good things about it. It's a pity we didn't see more integration and instead just saw everyone migrate away :(

Now, re: the rest of your post, AMD has issues with Barcelona and if they continue, we could well see barcelona based procs (at least on the higher-end, server side) pushed into next year.

No amount of sympathy will reduce a 2 year technology gap at a time when the company NEEDS to have a proper product on the market to compete and bring in revenue. This is pure, hard fact.

Further, Intel DOES have working silicon that happens to be quite fast and now they have increased their leadership on the mobile platform with Santa Rosa. Seeing as how there are more and more people going mobile, do you think AMD is going to be able to make any kind of dent with its higher TDP and higher power consuming parts vis-a-vis Intel?

AthlonXP1800
05-27-07, 02:58 AM
http://www.theinquirer.net/default.aspx?article=39756

Wow that is a massive blow to AMD, Barcelona is nothing what AMD claimed to be 40% faster.

AMD demorated 16 cores Barcelona vs 8 cores Opteron running POV benchmark with same settings on both systems on Tuesday, Barcelona scored 4000 pixels with 16 cores (4x4) and Opteron scored 2200 pixels with 8 cores (2x4).

I dont understand why AMD wanted to demorated 16 cores Barcelona vs 8 cores Opteron, that not very fair benchmark.

What happened if AMD did showed fair benchmark with Opteron, Barcelona and Clowertown, the estimates would looked something like below for 16 cores, 8 cores and a single 4 cores CPU:

4x4 16 cores

Opteron: 4400 pixels
Barcelona: 4000 pixels
Clowertown: 9800 pixels

2x4 8 cores

Opteron: 2200 pixels
Barcelona: 2000 pixels
Clowertown: 4900 pixels

1x4 4 cores

Opteron: 1100 pixels
Barcelona: 1000 pixels
Clowertown: 2450 pixels

Barcelona is now look like a joke, it will not save AMD as both R600 and Barcelona is killing AMD. Here is nothing to save AMD, it look more and more like that AMD is now finished, they continue lose marketshare and money, they cant compete with Clowertown, Penryn, Yorkfield and Nehalem.

I probably will wait for quad core Penryn later this year with 12MB cache or wait next year for Nehalem.

Sazar
05-27-07, 03:02 AM
:headexplode: