Conroe @ 2.66Ghz vs FX-57 @ 3.6Ghz: 3DMark05 CPU Test

[freecableguy]

freecableguy ***xtremesystems.org exclusive information***

Benchmark: 3DMark05 CPU Tests - single threaded application
AMD FX-57 @ 3.6Ghz vs. Intel "Conroe" ES @ 2.66Ghz

CPU: AMD 0516 FX-57 @ 3.6Ghz (257x14) @ 1.7v
Memory: 2x512MB PC-3200 Mushkin BH-5 @ 2-2-2-5, 257Mhz 1:1
Motherboard: DFI LanParty Venus NF4 SLI-DR
Cooling: Custom 'Chilly1' single-stage autocascade
CPU Temps: -30C idle, -15C full load while benching
OS: WinXP Pro SP2

CPU Test 1: 4.0 fps (compared to Conroe: -35%)
CPU Test 2: 7.4 fps (compared to Conroe: -13.5%)

FX-57 @ 3.6Ghz score: 8146 (compared to Conroe: -23%)




CPU: Intel "Conroe" 2.13Ghz ES @ 2.66Ghz (333x8) @ 1.15v
Memory: 2x512MB PC2-4200 Mushkin @ 3-2-2-4, 333Mhz 1:1
Motherboard: Intel D975XBX "Bad Axe"
Cooling: Stock Intel aluminum heatsink, ambient ~22C
CPU Temps: 35C idle, 41C full load while benching (according to EVEREST)
OS: WinXP Pro SP2

CPU Test 1: 5.4 fps (increase over FX-57: +35%)
CPU Test 2: 8.4 fps (increase over FX-57: +13.5%)

Conroe @ 2.66Ghz score: 10038 (increase over FX-57: +23%)




-FCG xs.org

[freecableguy]

Thanks to Absolute_0 for the following benchmark for comparison:

CPU: AMD Opteron 170 @ 2.66Ghz (266x10) @ 1.325v
Memory: 2x1GB ??? @ 3-4-3-8-1T, 266Mhz 1:1
Motherboard: ???
Cooling: water cooling
CPU Temps: not reported
OS: WinXP Pro SP2

CPU Test 1: 4.2 fps (compared to Conroe: -28.5%)
CPU Test 2: 7.4 fps (compared to Conroe: -13.5%)

Opteron 170 @ 2.66Ghz score: 8347 (compared to Conroe: -20%)

[Supertim0r]

really nice
me want one

[Absolute_0]

No way... get out of town!

What can i say, a very big
for FCG

Amazing results... just great. Very professional and well documented benchmark too, i like the blue and green color coding. Great Job Kris.

I know this was some work, but can we look forward to this kind of comparison for other benchmarks too?

[iterations]

Damn.... it really does look like clock for clock it is about 40% better. I seriously hope the haters don't give you are hard time about the graphics cards being different. Here is a preemptive: "For god's sake, this is the CPU test, the graphics cards don't matter!".

Cheers, FCG, keep it up!

[vapb400]

very very nice comparison. I hope you will do more of this style in the future.

[integral]

[bachus_anonym]

Too bad you don't have AMD X2 to compare with, though
From what we've seen on the ORB, X2 needs ~3.0GHz to match 5.4fps/8.4fps... That would give Conroe healthy 300-350Mhz as far as 3D benchmarking goes.

[Burdman]

Very nice FGC, glad to see some user-end comparisons... like others have said, it would be great if you could do some other benches, but I know you did a lot of work for what you gave us!

Thanks man, looking forward to more results.

[Absolute_0]

Too bad you don't have AMD X2 to compare with, though
From what we've seen on the ORB, X2 needs ~3.0GHz to match 5.4fps/8.4fps...

Sounds about right, 05 benefits a decent bit from dual core. Still 2.66 Conroe @ stock volts and stock heatsink beating 3 Ghz X2? sounds good to me. And beats 3.6 Ghz FX57!! lol, single core is out

[Burdman]

Sounds about right, 05 benefits a decent bit from dual core. Still 2.66 Conroe @ stock volts and stock heatsink beating 3 Ghz X2? sounds good to me. And beats 3.6 Ghz FX57!! lol, single core is out

BTW it's a 2.13 conroe oc'd to 2.66... not quite stock, but it still makes me happy

[Absolute_0]

BTW it's a 2.13 conroe oc'd to 2.66... not quite stock, but it still makes me happy

Yes i know, but stock volts and stock heatsink... and the overclock is still restricted by the board! It's comparable to X2 @ 3 Ghz, i'd like to see that on stock volts with stock heatsink!

Just my speculation, i'll try not to get too carried away. I just absolutely cannot wait to see what Conroe 3.33 Ghz does, and how far it clocks

[VulgarHandle]

look, i'm loving what conroe is showing, but putting a single core against a dual core isn't fair, single-threaded application or not, and there was no need to have seperate vid cards, as you could just use 1 card, and use the same in each setup, and if possible, use an opty dual-core, or an x2 with 2x1mb L2, and of course, run them both at same clock speed

what would also be nice is if you used 2x512mb for both, and find some tccd that can run 333

again, mucho appreciation for all your work, and again, i know conroe is powerful, but leave nothing for the flame-o's to feed on

edit: this would be the best possible scenario to show what cpu is better

[freecableguy]

3.6Ghz vs. 2.66Ghz isn't fair either....there is no fair.....let it go.

[OrcHuman]

Lol awsome!!!!!!! .You know what would be funny to see is a non FX cpu thats dual core and clocked to 2.6. I wonder how much faster that conroe will be HAHA! 60%? :P
Great work man!

[ORCBEAST]

3.6Ghz vs. 2.66Ghz is fair either....there is no fair.....let it go.

haha man i have been here since April 1st and I have seen that line pop up about million times....

Getting a HEADACHE ! haha Awesome scores..."off to make another thread in EOCF :P "

[Jamo]

thanks for the comparison freecableguy, looks to have made up my mind about my next upgrade being conroe based

[Miwo]

Great post FCG,

However, 3DMark05 does use multithreading in the CPU Tests. I am pretty sure vertex shaders are done with both processors, so you are really comparing Single core vs Dual Core. Conroe is definately an awesome chip though.

I just looked back at my own Dothan vs Yonah results, my previous Dothan @ 2.65GHz gets about 4700 CPUMarks, whereas my current Yonah setup gets 8874 CPUMarks. Sure, I may have different ram, mobo, etc..... but that big of a difference in my scores does not come from the new architecture alone, but rather because its dualcore and it does make use of it in the benchmark.

Another link showing Dualcore vs Singlecore 3Dmark05
http://www.xbitlabs.com/articles/cpu...x2-3800_5.html

[kiwi]

Would be good to add Yonah and X2 @ ~3Ghz just for information

[wittekakker]

the difference is just huge

[Absolute_0]

Alright i've run a comparison of my own

AMD Opteron 170 @ 2.66 Ghz
10x266 1:1 2gb ram @ 3-4-3-8-1T

3dmark05 CPU score = 8347 (approx. 17% slower than Conroe)
CPU test 1: 4.2 FPS
CPU test 2: 7.4 FPS

[Miwo]

Exactly my point...
This is a dual-core vs Single-Core comparison :o

[Repoman]

Good comparison absolute.

So the result isn't THAT big a deal after all.. clock for clock, less than 20% on this test. Don't get me wrong, that's awesome, but it wasn't fair to test against a single core apparently

[wittekakker]

Well it's pretty obvius the cpu scores are higher because it works kinda like 3ds max rendering. Dual core will always have a big lead there. I'm more interested how much difference there is on the final score with same GFX and software etc etc.

[misteroadster]

Thx absolute and now at 3ghz ?