AMD Fanboys start crying:

Big boys dont cry!

*hmmm and this is coming from a site ive never heard of, rofl!!! it doesnt include any actual screenshots of the benchmarks either. ha!

At first I was like, oh geez, here's another one of these dog eat dog threads. After reading this article though, I found it interesting and still do to this day that a mobile chip can run on par and sometimes faster than the fastest desktop chips.

I'm not sure who's crying at the moment. (and the tone that came across in your post does not give good impressions, I'll just note) To be honest, I'm not surprised to see the results whatsoever. Intel's Banias/Dothan/Conroe/Yonah arcitecture has always shown to be vastly superior to Netburst (used in the Pentium 4) and, at equal clock speeds, equal and often a tad better than the K7/K8. (Athlon64)

Something that did suprise me, though, is the Core's performance at some apps I lways considered "linear," and would hence provide an advantage to the high-speed Pentium 4/D. Of course, multi-threading was likely the culprit of the performance results for these things...

I personally appluad Intel for literally pulling this rather excellent arcitecture right out of their pocket. I can only wonder how it might further be enhanced when they include all of the other benefits found in competing chips, such as finding Hyper-Threading enabled, (which should show up in the first Core Duo Extreme Edition this fall, I think) as well as perhaps finally moving to an integrated memory controller as per the Athlon64/64 X2 series.

As for the "Conroe" chips, I'm really feeling dubious about where the better performance is going to come from. I'm suspecting that it may simply be higher clock speeds, as there have only been claims from intel on performance, not necessarily performance-per-clock-cycle. Elsewise, I'm thinking it may simply be the inclusion of Hyper-Threading as a standard feature in dual-cores, instead of just Extreme Edition chips. At either rate, Intels' got some chips to look forward to, and let's hope they can get them out quickly enough, and at a reasonable enough price, and they'll have their shot back at AMD. 'Bout time, too, as it's been some 2 1/2 years.

Who the **** cares.

they r selling computers with different configurations. best of their game pc is based on Duo but their workstation is based on FX.
http://www.gamepc.com/index.asp

I think those benchmark is true.

I don't care nor do I give a F***.

Interesting article...no real surprise tho given that Intel is now using their mobile architecture in their mainstream/desktop chips. I'd really like to get one of those A-Open boards and a Dothan or Yonah and build a SFF machines or HTPC.

But why title it "AMD Fanbots start crying"?!?! Are you trying to start another Intel/AMD flame thread? Got nothing better to do? Come on, don't be a wanker...

Literally the bottom line of your article is:

quote
".....and we haven’t been truly impressed with an Intel product in several years.
end quote

http://hkepc.com/hwdb/yonah-crossfire-3.htm
Trustworthy^^

amd boyz...rough days ahead for you......you can cry now....

Intel says conroe can do 4 operations per clock when Core duo can do only 3. They were the 20% advantage and remember Core Duo was laptop processor with slow FSB.
Put a workstation with two independent FSB and two dual core Conroe At more than 3Ghz and Bye Bye current processors ( Pentium D / AMD K8).

Look at intel latest roadmap.

July-August ---- Server ( were Opteron is now the winner and they are losing market)

October-November --- Desktop ( ready to school) (Apple will like it in educative enviroment)

December ---- (Laptop) ( they dont need 64bits before vista) ( no laptops with more than 4Gb)

Too bad they didnt include the Pentium EE 965 in the test.

Yeah, it is dead end. :wink:

What I think is interesting is the performance numbers from the Core Duo processors. Most users seem to agree that these are great chips and an excellent design from Intel. There are even a large number of AMD die hard users that agree with the above statement. So why is it so hard to believe that the Conroe processors will not perform well also?

i have other stuff to worry about

- like school




:wink:

Ah, yes, I had forgotten that the actual Conroe chips will bring about 1066MHz as the standard FSB speed, finally leaving that old 800MHz FSB behind. However, I have had poor luck finding exact information on the operational capacity of any particular Intel chips, and had presumed that for ages, they've been all capable of one integer and one floating-point operation per clock cycle.

As for Intel's new chips, the real challenge in fighting AMD will be to get to the battlefield before they leave. Remember, we're comparing chips that aren't out yet to chips that have been out for months. A 2.66GHz Conroe Extreme Edition sounds might impressive today, but will be meaningless if AMD pushes ahead, and perhaps with their first 65nm chips, produces a 3.0GHz, or even 3.2GHz, Athlon64 X2 (FX-64, anyone?) that's there at the same time.

It is a bumpy road ahead, but if Intel doesn't slow down, they've got a shot to catch up or even get ahead. Their first test will be with the servers, as you note; to say that Intel's "lost market share" to the Opteron would be an understatement. Intel's chips (as well as Dell's refusal to use anything but) have become the brunt of countless jokes in the server market. It's also an embarrassment that Intel chips are only found in a handful of modern supercomputers... And then, only because the chip's ludicrously poor sales caused a precipitous decline in their price, making hooking together 131,072 of them a financially feasible idea. At any rate, I doubt Intel can possibly win this fight, at least this year, but they may be able to staunch their already existing wounds.

Of course, the main fight we're looking at is desktops. Conroe promises a lot, but as I noted, if it winds up being too little, too late, they could lose their shot. Particularly with AMD already going to have a few months' head start with AM2. All that's left to be seen is the lineup AMD will bring. Intel best cross their fingers for another delay bringing 65nm to the table, as that's one of the biggest cards Intel's got to play.

Ironically, the last battle is the easiest. I think Intel annoyed a lot of people by announcing Merom for the END of the year. However, business-wise, it makes sense; it's the least pressing front, as they've already got popular support, and AMD's chips, in spite of their technical merits, simply don't match Intel's here.

WOW, that is indeed interesting.
More so, the power output, and the noise factor. As a network engineer, my home lab is REDICULOUS. when it comes to POWER + HEAT + Noise.

Now intel just needs to make the prices more attractive. $700+ for a mobile cpu is kinda steap.. The speed is nice though, i have to admit.

Does anyone else feel like this, that we are just like hovering in mid air, until Vista comes out. I mean, no direct x 10 vid cards out, no sound. Both amd and intel comming out with new cpu's in like 2-3 months. I dont even want to look at another computer price list, until the dust settles..

so basicaly were just going to get into a clock race between amd and intel again?

because the clock vs clock looks pretty much even to me

What I find most humorous is the fact that intel fanboys like our parent poster here have been more than willing to take it in the ass from intel for YEARS. Patiently bent over, waiting for their reach-around, and, now that they actually get it, are grateful for the whole experience.

My eyes are blue

I haven't been in Intel's camp for the past 4 years or so, but this Core Duo firmly puts me in it, at least mobility wise. Core Duo is an impressive piece of engineering given Intel's recent history of messed up sh!t. Mobile platforms have been evolving at a much better pace than the desktop variants imo.

What I really want to see is a DFI based board (or something to that effect) for Yonah just to see how much you can trully get out of the proc. AOpen makes decent boards, but when it comes to overclocking its hard for anyone to argue that DFI flat out makes the best boards. Given they are complicated and tempermental but they are crazy fast.

I am looking forward to Conroe, but as you said the challenge is getting it to market. Its unfair to compare a Conroe to a Toledo. If Conroe didn't beat a Toledo then I would be severely concerned as to why Intel would release such a chip. Each generation is meant to be faster than the previous. I am sick and tired of all the Intel fan boys waking up and immediately saying "OOO AMD IS DEAD." I realise alot of the die-hard AMD fanboys really give the entire AMD camp a bad name, and I don't like those guys anymore than I like the snobby Intel fan boys that have awoken.

I am impressed with Intel IF they can really bring Conroe to market with that 20% improvement that they are claiming. I am also impressed with AMD's ability to shoot for a 65w TPD on 90nm chips. In the end we all win.

OK, I shouldn't, but I'll bite.

I suppose that some might label me an AMD fanboy, I don't think I am, but I suppose one could draw that conclusion. I do prefer AMD over intel, but I've purchased both, and am currently running active PC's with both. My main one is a new X2-4400 build.

With that said,...I'm not sure why I'd be crying.....especially over a chip that I can't buy yet....and won't be able to buy for quite some time.

Should I be crying because intel hasn't come out with it yet?

I suppose I need some help telling my tear ducts why they should waste some tears....cause I don't get why that link shouuld make me cry.

So obviously an Intel Fanboy, to even post this thread. Proof that he hasn't even read any previous threads, since the AMD/Intel debate has been going on since the cavemen.

Let me give the ultimate breakdown:
AMD and Intel will be going back and forth for the immediate and long-term future over who is the performance champ. Prices on the chips will also fluctuate back and forth for the immediate and long term future over who has the best value. The only winner is the consumer, to whom these companies must compete for their money.

Now can't we all just get along?

some of you have false understanding of whats going on, and that is where I think AMD is enjoying some time to relax. First off, AMD won't have anything NEW coming other than the socket till Q1 or even Q2 of 2007. That is when their 65nm K8L will come out. But again, AMD doesn't need to switch to 65nm to compete with Intel. The reason Intel HAS to go to 65nm is because of the cache and how their archetecture works. Amd doesn't HAVE to move to 65nm to compete. We all know that. heck Intel is going to be on 45nm around the same time as AMD hits 65nm but AMD won't hit 45nm till 2008, but they don't NEED to. They have a much more effecient die. AM2 is just a socket, and a revision to an aging, yet powerful architecture.

Intel WILL have a better product out sooner than AMD, but will it be NECCESSARY? Thats the real question. Most people, the market, is driven by OEM system builders who choose either AMD or Intel. AMD has contracts with these builders that extend past Intel's Conroe launch. Yeah for the enthusiasts that building there systems, there will definately be some that leave AMD's camp, but I really doubt it will be any of the people that are now enjoying thier 3800+ or X2's...the upgrade path of those people will most likely fall again once AMD has their new chips out next year and also when people start to upgrade for Vista with new DX10 cards, and such. The server market is still most likely going to be on AMD's side, for some of the same reason. A lot of servers were bought in the past couple of years from AMD because of the large performance difference they had over the Itaniums and such. Corporations do not like to upgrade servers THAT much, so when will they want to? When AMD releases their new chips, which SHOULD be on par or better than Intels solution at that time because they didn't JUMP THE GUN like Intel loves to do.

It's all speculation, but I would say that Intel will not see a GREAT increase in profits from Conroe. I think its great that they were able to finally come back at AMD, but I think it was too late. Too many people already bought new systems in this past few years that didn't include a Intel chip. I do however think they will certainly see a good number of Core Duo systems sold as they seem to be doing quite well, but again corporations seem to love those silly IBM machines which are not running Intel inside.

No, Conroe is more efficient per clock than K8. Remember the 2.66 Conroe beating a 2.8 FX60? Yeah. 65nm is something AMD needs to compete in performance.

Hmm something doesnt seem right if yonah can do 3 operations per clock and amd k8 can do the same then how come at the same speed its beating it in F.E.AR.? Could it be ddr2 or could it be that its an intel chipset with an intel cpu has less bottleneck while amd has 3rd party chipsets from nvidia hmm? Look at F.E.A.R. and Half Life 2 Lost Coast it just doesnt seem right the ram timings on yonah(core duo) are higher and they get better fps?

Hmm something doesnt seem right if yonah can do 3 operations per clock and amd k8 can do the same then how come at the same speed its beating it in F.E.AR.? Could it be ddr2 or could it be that its an intel chipset with an intel cpu has less bottleneck while amd has 3rd party chipsets from nvidia hmm? Look at F.E.A.R. and Half Life 2 Lost Coast it just doesnt seem right the ram timings on yonah(core duo) are higher and they get better fps?
Like I said, Intel P'M architecture is more efficient per clock than K8, unlike Netburst. Also, the more frequency you add to RAM, the less latency it has despite timings. If you expect more than 5% increase with DDR2 on AM2, prepare to be disappointed.

I'm not so sure that AMD needs 65nm. Which btw who decided on 65nm and not 62nm? just a thought. But anyways, AMD is fully capable of matching Conroe on 90nm if they wanted to.

Think about it... how long has the toledo/manchester architecture been in prodcution? And Conroe isn't even out yet, so again I would expect Conroe to beat Toledo cores. What we all should wait for is AM2 v Conroe since those are both new chips.

AMD has been beating Intel's 65nm for awhile (margin decreased with presler but AMD still wins), only with Conroe will Intel's 65nm process finally overtake AMD.

Hmm something doesnt seem right if yonah can do 3 operations per clock and amd k8 can do the same then how come at the same speed its beating it in F.E.AR.? Could it be ddr2 or could it be that its an intel chipset with an intel cpu has less bottleneck while amd has 3rd party chipsets from nvidia hmm? Look at F.E.A.R. and Half Life 2 Lost Coast it just doesnt seem right the ram timings on yonah(core duo) are higher and they get better fps?
Like I said, Intel P'M architecture is more efficient per clock than K8, unlike Netburst. Also, the more frequency you add to RAM, the less latency it has despite timings. If you expect more than 5% increase with DDR2 on AM2, prepare to be disappointed.I think i've figured it out core duo uses ddr2-667 with timings of 5,5,5,10 and core duo's fsb is 667 then the ram runs at 1:1 ratio while the fx-60 runs at ddr 400 with 2.5,3,3,7 and the fx-60 fsb is 1000 each way(upstream and downstream HT Link) and since ddr2 ram is like an 8 lane highway(4 up 4 down)(and they were running dual channel so the ram acted like a 16 lane super highway)and since ddr ram is like a 4 lane freeway(2 up 2 down)(also dual channel but 8 lane highway) could it be that the ram is responsible for these results? I think if they retested the core duo with ddr ram the results would be almost even with the fx-60 dont you think so?

I think i've figured it out core duo uses ddr2-667 with timings of 5,5,5,10 and core duo's fsb is 667 then the ram runs at 1:1 ratio while the fx-60 runs at ddr 400 with 2.5,3,3,7 and the fx-60 fsb is 1000 each way(upstream and downstream HT Link) and since ddr2 ram is like an 8 lane highway(4 up 4 down)(and they were running dual channel so the ram acted like a 16 lane super high way)and since ddr ram is like a 4 lane freeway(2 up 2 down)(also dual channel but 8 lane freeway) could it be that the ram is responsible for these results? I think if they retested the core duo wqith ddr ram the results would be the same dont you think so?

I don't think the RAM is having that big of an effect on performance. It will effect it, yes but with the much higher latencies of the DDR2 the 667 speed is necessary to keep pace with the DDR. I am wondering why they ran a FX 60 with 2.5-3-3-7 timings and not 2-2-2-5 timings, hmmm.

Just wondering, can I use a socket 478 Core Duo Yonah on my old i865 socket 478 motherboard?
It will be VERY cool if I can. !! Please write back about this...

ummm did anyone notice:

This lab report was updated on 4/20/2006.
With the removal of two jumpers, this AOpen motherboard does allow FSB speeds up to 1066 MHz FSB. Using this technique, we were able to overclock a Core Duo T2600 (2.16 GHz) to 2.9 GHz.

That's almost a 1 Ghz overclock... can't wait to see what a conroe can do at 3 Ghz + .

I think i've figured it out core duo uses ddr2-667 with timings of 5,5,5,10 and core duo's fsb is 667 then the ram runs at 1:1 ratio while the fx-60 runs at ddr 400 with 2.5,3,3,7 and the fx-60 fsb is 1000 each way(upstream and downstream HT Link) and since ddr2 ram is like an 8 lane highway(4 up 4 down)(and they were running dual channel so the ram acted like a 16 lane super high way)and since ddr ram is like a 4 lane freeway(2 up 2 down)(also dual channel but 8 lane freeway) could it be that the ram is responsible for these results? I think if they retested the core duo wqith ddr ram the results would be the same dont you think so?

I don't think the RAM is having that big of an effect on performance. It will effect it, yes but with the much higher latencies of the DDR2 the 667 speed is necessary to keep pace with the DDR. I am wondering why they ran a FX 60 with 2.5-3-3-7 timings and not 2-2-2-5 timings, hmmm. Yeah that is odd isn't it what if they used this ram it has the tightest timings i could find for a dual channel 2x1gig kit for the amd platform and this ram for the intel platform also tightest timings i could find

No, just No.

I think i've figured it out core duo uses ddr2-667 with timings of 5,5,5,10 and core duo's fsb is 667 then the ram runs at 1:1 ratio while the fx-60 runs at ddr 400 with 2.5,3,3,7 and the fx-60 fsb is 1000 each way(upstream and downstream HT Link) and since ddr2 ram is like an 8 lane highway(4 up 4 down)(and they were running dual channel so the ram acted like a 16 lane super high way)and since ddr ram is like a 4 lane freeway(2 up 2 down)(also dual channel but 8 lane freeway) could it be that the ram is responsible for these results? I think if they retested the core duo wqith ddr ram the results would be the same dont you think so?

I don't think the RAM is having that big of an effect on performance. It will effect it, yes but with the much higher latencies of the DDR2 the 667 speed is necessary to keep pace with the DDR. I am wondering why they ran a FX 60 with 2.5-3-3-7 timings and not 2-2-2-5 timings, hmmm. Yeah that is odd isn't it what if they used this ram it has the tightest timings i could find for a dual channel 2x1gig kit for the amd platform and this ram for the intel platform also tightest timings i could find
They use that RAM in every test. Crappy ram was used for both platforms, in NO way was the test rigged. BTW, the fastest 2x1gb set for DDR1 is the Corsair 3500@2-3-2-5.

I meant at ddr 400 speed

Regardless, memory makes little impact on overall performance.

Not to sound stupid, but aren't the first chips for AM2 going to be updated versions of the K8 architecture? Also, Conroe is pretty much an entirely new processor. Unless I'm really dumb and left out of the loop, why are we all comparing some mysterious benchmarks that may or may not exist (and even if they do, they sound fishy). Ok, even if we're not whining and "comparing" cpus to motherboards, seriously, you all need to get lives. Why would you care if Conroe is better? Or if AMD is better? Seriously, we win in the end either way. Don't you have anything better to be a fanboy about? Now, where is MadModMike to call me a spineless weasel? The whole lot of you (well most of you) are losers. My advice: get girlfriends/boyfriends. Talk to your friends. Hell, actually use your computer for something other than getting on these foums and bitching your fanboyisms into oblivion.

You have to understand that AMD was beating Intel on 130nm when Intel was on 90nm easily. Why? Because Intel NEEDED to use 90nm so that they could fit their core on the die along with all the cache they were shoving in there. AMD didn't Need 90nm until they switched to dual-core, and even then they were trumping them. Now Intel NEEDS to go down to 65nm to fit the extra cache again. Yes Conroe is more efficient than K8...K8 came out in 2003, not end of 2006 remember. Will Conroe be more efficient than a K8L at 65nm? who knows yet, but I can bet that it wont. And even then we are talking about a core technology basically made in 2003 and just updated with some fresh instruction sets and virtualization techniques.

As I said, I KNOW conroe will be crowned KING when it launches, but again, that won't matter. Just because you have the best product doesn't mean you sell more. If that was the case, AMD would have killed Intel starting a few years back. It takes time for companies to SEE the benefit of switching, it takes time for contracts to end too. No company that bought a AMD server in the past few years is gonna jump on woodcrest just because its the king of the hill for now. But once the upgrade path looks good for those companies, AMD will have its K8L out and we will see what will be more advantagous to them.

Enthusiasts don't drive the market. The market is driven by OEM system builders and server buyers. Right now I see NO reason for anyone to upgrade to conroe unless they finally have the money to spend on a high-end system. Laptop buyers usually buy cheap, but companies like to buy brand ie IBM.

For enthusiasts to pony up for a new MB, CPU, DDR2 ram it would take a lot for them to part with their 939 K8's, heck im even hesitant to part with my old 754 pin K8 64 3200+. I'm more likely to see what comes out in early 07 for both companies and decide then as I will need to buy a new MB, CPU, and RAM.

And to the guy that keeps saying AM2 is a new chip....its not. It's the same ol K8 with DDR2 memory controller. It will be better by about 10% but only with FAST DDR2 1066, and about 5% faster if paired with 667 ram, no more.

Aaarrggghhhh :x By next year my cpu will be outdated and outperformed and I'm going to have to buy new ones. Damn this AMD/Intel battle! :x Why can't they make a cpu that could last as long as a silicon modded breasts?!  

You have to understand that AMD was beating Intel on 130nm when Intel was on 90nm easily. Why? Because Intel NEEDED to use 90nm so that they could fit their core on the die along with all the cache they were shoving in there. AMD didn't Need 90nm until they switched to dual-core, and even then they were trumping them. Now Intel NEEDS to go down to 65nm to fit the extra cache again. Yes Conroe is more efficient than K8...K8 came out in 2003, not end of 2006 remember. Will Conroe be more efficient than a K8L at 65nm? who knows yet, but I can bet that it will. And even then we are talking about a core technology basically made in 2003 and just updated with some fresh instruction sets and virtualization techniques.

As I said, I KNOW conroe will be crowned KING when it launches, but again, that won't matter. Just because you have the best product doesn't mean you sell more. If that was the case, AMD would have killed Intel starting a few years back. It takes time for companies to SEE the benefit of switching, it takes time for contracts to end too. No company that bought a AMD server in the past few years is gonna jump on woodcrest just because its the king of the hill for now. But once the upgrade path looks good for those companies, AMD will have its K8L out and we will see what will be more advantagous to them.

Enthusiasts don't drive the market. The market is driven by OEM system builders and server buyers. Right now I see NO reason for anyone to upgrade to conroe unless they finally have the money to spend on a high-end system. Laptop buyers usually buy cheap, but companies like to buy rand ie IBM.

For enthusiasts to pony up for a new MB, CPU, DDR2 ram it would take a lot for them to part with their 939 K8's, heck im even hesitant to part with my old 754 pin K8 64 3200+. I'm more likely to see what comes out in early 07 for both companies and decide then as I will need to buy a new MB, CPU, and RAM.

And to the guy that keeps saying AM2 is a new chip....its not. It's the same ol K8 with DDR2 memory controller. It will be better by about 10% but only with FAST DDR2 1066, and about 5% faster if paired with 667 ram, no more.
Let's not forget AMD doesn't really come out with any important architectural changes until 2007 with K10. AM2 release is just a K8 with DDR2. The past doesn't really mean much here, and the reason is because Intel realized that Netburst lost efficiency per clock, which is now all important. So, obviously a 90nm Intel on a crappy architecture will lose out to a 130nm efficient K8. But to say that a 65nm highly efficient architecture will lose to a 90nm older series K8 is rediculous. Not only that, but Intel will make some headway provided they are ahead until 2007. Servers and laptops too, especially considering that servers want power management and low clock speeds Woodcrest has to offer, as do mobile users. July 2006 to the Q1 2007 is a long time in stock market years.

yeah i said that...And July 2006 to Q1 of 2007 isn't a long time when talking architecture changes. Netburst came out in 2002, K8 in 2003. Thats 4 years to a new "real change". So 2 quarters vs 12-16 quarters is nothing.

To those saying, well why wait something is always around the next corner, not when it comes to major architecture changes. Now is a really sticky time for both companies, Intel is gearing up to come out with a great product, but they are also spending millions trying to promote their new product before its even out, in essense telling people not to buy what they have out right now, forcing them to cut prices and cut revenue. AMD has been quite silent about the AM2 switch, but really what do they have to say. When joe smoe goes into a store, he doesn't ask if its running 939 or 754...he asks how cheap and how fast. Now Joe Smoe is going to be told by the salesman, hey buy this Core from Intel...but AMD has a core too lol. I think AMD put it well when they said, why would you market a car thats a chevy and call it a Engine....all cars have them. Anyhow, we shall see what happens again but as I said, contracts are contracts and AMD won't lose contracts to Intel over this. AMD might not gain anymore contracts after the launch until they answer in 07 but thats again not much time.

The average joe says,
I need ma Intel Inside bitch dont try to sell me this AMD trash!

I don't think AMD fanboys will be crying. However, I do think Yonah owners will be crying when they want to run 64-bit software on their 32-bit chip. I think Intel is out of their minds to release a 32-bit only chip in a world where even 64-bit single core chips are going out of style like Weird Al.

Sure, except Windows x64 is a f*cking joke.
Maybe with Vista though.

Hmmm.....ignorant words from an Intel Fanboy....hmmm....

Fact: Intel EM64T performs up to 20% below that of x86 Proof

Fact: If I owned a Yonah I would cry because I can't beat AMD in x64, nor can I in x86 without having 2MB or more of Cache

Fact: AMD64 performs up to 75% in 64-bit over EM64T and 90% over itself in x64 vs. x86 Proof

Hmm.....I should dry my eyes I guess

~~Mad Mod Mike, pimpin' the world 1 rig at a time

Well, it was a nightmare with little app and driver support but everythign that worked worked faster when I had it installed. I may reinstall it on my studio machine since I don't need all the apps I run on my gamingdev box.

Hmmm.....ignorant words from an Intel Fanboy....hmmm....

Fact: Intel EM64T performs up to 20% below that of x86 Proof

Fact: If I owned a Yonah I would cry because I can't beat AMD in x64, nor can I in x86 without having 2MB or more of Cache

Fact: AMD64 performs up to 75% in 64-bit over EM64T and 90% over itself in x64 vs. x86 Proof

Hmm.....I should dry my eyes I guess

~~Mad Mod Mike, pimpin' the world 1 rig at a time
**Notice no proof on #3. x64-86 emulation loses performance in many categories you might add, not to mention annoying compatibility issues.