- Back to homepage

Some very early benchmarks Conroe 2.4ghz vs 2.7Ghz A64 AM2

@ 2006/04/08

cpu @ 225*12=2700mhz (multiplier cannot be changed in the bios)
AMD's Stock cooler , vcore=1.32v~1.36v varies
ram @ 333mhz 5-5-5-5-15 1T 1GB*2 dualchannel
mobo = nv C51G
OS=xpsp2 with patch + nv 8.252 drivers

pi_1m=39.8sec(conroe 2.4G=21.25sec)
pcmark05 cputest=5510(conroe 2.4G=6101)
3dmark03 cputest=1073(conroe 2.4G=1413)
3Dmark05 cputest=5582(conroe 2.4G=8320)
scienmark2.0=1454(conroe 2.4G=1310)
pi_fast=46.03/58.44(conroe 2.4G=32.55/40.41)
sisoft alu=24220/11022
sisoft multi=51287/55489
sisoft mem=6919/6826
sisoft cache=7432

Comment from GIBSON @ 2006/04/09

oh well, indeed as hardfreak said, it's only a new package for an "old" cpu (kinda forgot about that) anyhow, i'm kinda curious what will happen when amd goes to 65nm

Comment from wutske @ 2006/04/09

AM2 cpu's, atm, are nothing more than current S939 cpu's with a new memory controler (and probably some smal tweaks here and there).
So, new technology is being compared with 'older' technology, in fact, we are comparing a cpu with 4 times the amount of L2 cache, wich is an advantage, not ?

I thing there might be a performance gain by using lower latency DDR2-800, on both systems, question is tough, how much of a gain (prolly not enough for AMD, because the Intel system used only DDR2-533, Low latency btw)

Comment from Rutar @ 2006/04/08

I doubt anyone thinks AM2 will not get completly owned by Conroe.

Comment from GIBSON @ 2006/04/08

he did mention he had some problems with the mobo and he did say "funny results", and besides, it's obvious that something is wrong somewhere
quite normal that something is wrong too this early, if it wouldn't have any faults it'd be already in production

Comment from Sidney @ 2006/04/08

Explanation we have had read, seen, experienced with A64 regarding latency in the past; doesn't change the result

Comment from GIBSON @ 2006/04/08

QUOTE:
"From what I gathered, it seems like Super Pi like a low latency cache/memory subsystem more than anything else, so you could probably argue that combining a small L2 X2 (512kB) with a rather high-latency piece of RAM is the worst case scenario for that software.

Then again, I guess we'll need more datapoints to come to anything definitive. (or at least some latency numbers for the AM2 sample so we can compare with current S939 X2s)"

that might be an explanation

Comment from The Senile Doctor @ 2006/04/08

definitive testing head to head in the holiday season

Comment from Sidney @ 2006/04/08

I have never expected DDR2 latency will produce good result by looking at so many of our reviews.
My Venice @2.7G produces 30" SuperPI and 7600mb in Sandra; Opty 165 @2.6, 32" and ~7000mb.

Comment from jmke @ 2006/04/08

AM2 won't offer large improvement over S939 so far;

Comment from Sidney @ 2006/04/08

Not exciting performance whatsoever

All information and graphics contained in Madshrimps are sole property of the Madshrimps crew and may not be reproduced or copied in any manner without written permission from us.