First Hammer benchmarks

[Kell]

http://www.tecchannel.de/hardware/937/0.html

Mostly just synthetic tests and a Q3A benchmark. If Q3A is anything to go by, the Hammer has one hell of a boost to IPC!

Could just be the memory subsystem though; Q3A is supposedly very dependent on that.

EDIT: corrected link to point to first page of article instead of last page.

 

[CedarTeeth135]

*see "amd concedes poor performance*

lock

 

[First]

Originally posted by: CedarTeeth135
*see "amd concedes poor performance*

lock

Kell is referring to something completely different than the story you're referring to CedarTeeth (you're referring to this story, Kell is referring to this story).

 

[SSXeon5]

Its sad that they compare a newer chip with a older Pentium 4. Hey if 800Mhz is there highest then they should compare it to a P4 2.53Ghz with some PC-1066 on a i850E

And if it does get released around dec then it wont be 2GHz at all, most defently like they said in the artical of 1.6GHz. At that point the 3Ghz P4 will be out and so will granite bay How high did that Clawhammer OC ... can it?

SSXeon

 

[BlvdKing]

I find it interesting that, at least in these benchmarks, a CH 800 MHz (256k L2) = P4 1600 MHz Williamette. That's a hell of a discrepency.

 

[Adul]

Originally posted by: SSXeon5
Its sad that they compare a newer chip with a older Pentium 4. Hey if 800Mhz is there highest then they should compare it to a P4 2.53Ghz with some PC-1066 on a i850E

And if it does get released around dec then it wont be 2GHz at all, most defently like they said in the artical of 1.6GHz. At that point the 3Ghz P4 will be out and so will granite bay How high did that Clawhammer OC ... can it?

SSXeon

why do i find you post juts a bunch of mumbling jiberrish. You are making statements that came strait out of your ass.

They had one hour with the chip, if you read the article.
They mention it was runing very cool to the touch.
It performance for even 800 mhz is very nice, though the limited benchmarks give a norrow focus.
This is an unofficial bench, one early hardware, drivers, chipsets, etc.
Wait till it is release and stop acting like a fool please. I can appreciate you enthusiam, but stop making comments that you can not prove squat. We all have to wait and see what will be released and at what speeds.

I will say it will be a competitive chip.

 

[mamisano]

Well, when this processor finally gets up to speed, it is going to be REAL fast.

One question. How much of the performance increase in Q3 over an MP@800 comes from SSE2?

 

[Adul]

Originally posted by: mamisano
Well, when this processor finally gets up to speed, it is going to be REAL fast.

One question. How much of the performance increase in Q3 over an MP@800 comes from SSE2?

i dont think Q3 supports SSE2, just SSE

 

[mamisano]

Originally posted by: Adul

Originally posted by: mamisano
Well, when this processor finally gets up to speed, it is going to be REAL fast.

One question. How much of the performance increase in Q3 over an MP@800 comes from SSE2?

i dont think Q3 supports SSE2, just SSE

I am just getting TOTALLY confused with all the new products coming out

 

[Kell]

Originally posted by: SSXeon5
Its sad that they compare a newer chip with a older Pentium 4. Hey if 800Mhz is there highest then they should compare it to a P4 2.53Ghz with some PC-1066 on a i850E And if it does get released around dec then it wont be 2GHz at all, most defently like they said in the artical of 1.6GHz. At that point the 3Ghz P4 will be out and so will granite bay How high did that Clawhammer OC ... can it?

The Hammer is pre-release. This was mainly an IPC test and memory subsystem test. You already know that released Hammer samples are locked at 800MHz; no OC allowed there, so why are you even asking?

No need to get all up in arms about this or about Hammer. It was just a friendly posting of benchmarks that many people (including you) wanted to see. I know we'd all like to see it at higher MHz, but this is all (actually more) that AMD has chosen to let the general public see.

 

[imgod2u]

I think this has got to be the most posted-about topic to date in such a short time. I think there have been what, 10 different posts on both here and the general hardware forum within the past day or two.

That said, if this is any indication of how well Hammer will perform it is impressive to say the least. And it does sound about right. At 2.0 GHz it would probably be similar or even edge out a 3.0 GHz P4 with a 133MHz QDR FSB.

 

[SSXeon5]

Originally posted by: Adul

Originally posted by: SSXeon5
Its sad that they compare a newer chip with a older Pentium 4. Hey if 800Mhz is there highest then they should compare it to a P4 2.53Ghz with some PC-1066 on a i850E

And if it does get released around dec then it wont be 2GHz at all, most defently like they said in the artical of 1.6GHz. At that point the 3Ghz P4 will be out and so will granite bay How high did that Clawhammer OC ... can it?

SSXeon

why do i find you post juts a bunch of mumbling jiberrish. You are making statements that came strait out of your ass.

They had one hour with the chip, if you read the article.
They mention it was runing very cool to the touch.
It performance for even 800 mhz is very nice, though the limited benchmarks give a norrow focus.
This is an unofficial bench, one early hardware, drivers, chipsets, etc.
Wait till it is release and stop acting like a fool please. I can appreciate you enthusiam, but stop making comments that you can not prove squat. We all have to wait and see what will be released and at what speeds.

I will say it will be a competitive chip.

hmm ... What "statements" did I make? O you mean when I said "most defently" .... Im sickened that its been about 5 monthes and after Amd said that there chips is so fast because its a K8 .... and its still at 800Mhz .... does it take them 5 monthes+ to get new steppings or silicon? I mean first off they compare a Clawhammer with little info on the test setup and a Willy P4 lol. Wheres the northwood with PC-1066 and 533Mhz fsb? Or some DDR333 and a sisDX?

I can appreciate you enthusiam, but stop making comments that you can not prove squat.

What Am I proving? All I said was they should bench flagship chips ... not old willys. And from some of the reports that if it releases in Q4-2002 it will be around 1.6GHz, I never stated concrete facts. Yet I get bashed for freely expressing my opinion .... 1st amendment anyone?

SSXeon

 

[Kell]

Originally posted by: SSXeon5
hmm ... What "statements" did I make?

Let's see, you state that:

1) If Hammer does get released by December, it won't be 2GHz at all.

2) "At that point" (December?) the 3Ghz P4 will be out and so will granite bay.

While you might not consider them "definite statements," such comments could sound very definite to the average layman, or even the average computer geek. Right now neither "statement" is certain.

Originally posted by: SSXeon5
Amd said that there chips is so fast because its a K8 .... and its still at 800Mhz .... does it take them 5 monthes+ to get new steppings or silicon?

Actually, no. They've got the Hammer farther than 800MHz, and that is pretty definite. AMD's just not giving the public anything above 800MHz, nor are they giving anyone I know of overclockable samples.

Originally posted by: SSXeon5
I mean first off they compare a Clawhammer with little info on the test setup and a Willy P4 lol. Wheres the northwood with PC-1066 and 533Mhz fsb? Or some DDR333 and a sisDX?

It was an IPC test more than anything, and they only had an hour with the thing. A full-blown review like you're expecting often takes several days or even weeks.

Originally posted by: SSXeon5
1st amendment anyone?

First amendment cuts both ways. Everyone has the right to speculate, make statements, and (within limits) even outright lie (please note i'm not accusing you of lying). Everyone also has the right to contest, debate, and (within limits) even outright flame.

 

[broadwayblue]

what will the fsb be on the hammer? i assume it will be increased from the current 266, right?

 

[JeremiahTheGreat]

wasn't the FSB on the lines of 800Mhz?

 

[andreasl]

The "FSB" speed on Hammer is pretty irrelevant to it's performance in single CPU setups. It just have to be fast enough to handle the AGP and other IO traffic. This is because of the on-die memory controller. It means that the memory data does not travel across the Hammer FSB (HT links) except in dual+ setups.