Printable View
A couple of benchmarks I've done with these two cpus.
Benched with the opteron, removed it and benched with the X2 with exactly the same settings.
http://img485.imageshack.us/img485/2...chmarks5gu.gif
ps: just recently I've done the same with a opteron dual core 165@2750 but used different drivers so I'm not posting them.
I'd be interested in the dual core Opteron results.
great job - nice to see the advantages of dual core are enough to practically match in most cases if not beat a single core at 400mhz lower clock. FX-60 is gonna rock the world :D
ive been doing similar tests with my X2 4200 against my 146
X2 scores slightly more @ 2.7ghz than my 146 does @ 3ghz so the extra L2 doesnt seem to be making much difference, about 100mhz looking at your results :toast:
why 11x237 and not 10x260 ?
The patriot's 3200llk 2gb kit that I've used wouldn't do 260, tops at 258.
why? he tested a toledo core, which is 1mb L2, just like the dc opteronQuote:
Originally Posted by Zim
I'm sorta skeptical of the results. The Doom3 engine is multi-threaded, but not the others. There has been countless review site benchmarks of X2 vs A64. Clock for clock with single threaded games, the X2 would either Tie, or beat the single core by 2-3% tops.
Unless the Opteron is slower somehow..
nvidia's drivers enhance for 3d performance whether or not it was optimized for dual core
this was ever since the 8xx.xx series drivers
Benchmarks are 100% fair.
No other program or devious process was running in background. Pc was fresh installed with xpsp2+patches+drivers. Opteron was tested with the default sp hal and then X2 with the default dp hal.
No amd x2 drivers or hotfixes were installed.
I think it's all related to nvidia driver optimization in 80x series.
If you see the latest tom's hardware "The Mother of All CPU Charts 2005/2006" using older 78.01 drivers, a stock FX57 tops almost every game benchmark.
http://www.tomshardware.com/cpu/2005...charts-23.html
I for one lost 400 in 3dmark05 by running my opteron at 2.8 vs my X2 3800 at 2.6, so these numbers look pretty good to me. Hats off to nivida.
the drivers don't make a game multithreaded, just remove a lotta the driver overhead to the other core, which apparently is pretty decent amount. the engine is still on one core and i'm pretty sure doom 3 isn't multicore supporting in and of itself(as every single benchmark dual core to none dual core showed clock for clock ath 64 vs ath x2 performance. quake 4 however seems to have been modified properly to dual(maybe multicore)core as when you watch task manager both cores share the load pretty evenly and tend to stay at similar percentages. doom3(from what i remember, and from reviews) shows a large amount done on one core and the driver overhead doing a decent amount of work on other core, but not anywhere near even.
http://www.xbitlabs.com/articles/cpu...u-games_6.html
Can you install the drivers and hotfixes and not do the registry entry on the hotfix and then bench and see what you get??Quote:
Originally Posted by japhex
I am interested to see if the hotfix or the driver have any affect on performance.
These results are precisely why I wouldnt suggest a single core anything to anyone riht now. And to think I got grief from folks for ditchan fx55 for a xs.. Geez