Test to confirm the AMD-60 numbers from the Conroe benchmark

[Diasper]

Reading the comments for the Conroe preview benchmarks there's a lot of people questioning the validity of the results. There's a simple way of sorting this.

Does anyone have dual-core with 2MB cache overclocked to 2.8ghz with 1900XT crossfire to benchmark the FEAR demo?

Good idea/bad idea?

 

[kyparrish]

I'd like to see this as well.

Someone did point out in the comments though that since FEAR is so shader-intensive, that the crossfire x1900's might just actually be showing that FEAR has been cpu limited.

Wow, I really can't wait to see how this all plays out.

 

[fierydemise]

From the AT X1900 review with a stock FX-57 (2.8Ghz) they got 133fps and in the conroe article they got 132fps, both are clocked at 2.8Ghz and FEAR isn't multithreaded the performance of both CPUs should be the same and thats what the results indicate so I don't see anything fishy with the results.

 

[kknd1967]

en how about this:
FYI

different resolution though

 

[fierydemise]

Originally posted by: kknd1967
en how about this:
THG

those are 1600x1200 the AT numbers are 1240x960

 

[Diasper]

From what I've managed to gleen the numbers look approximately correct (through likely extrapolation) BUT it's hard to say with certainty for a number of factors.

Primarily, as an immature tech Crossfire is still seeing significant performance improvements with driver releases and so to use earlier reviews may not be accounting for more recent versions of it.

It'd be nice therefore to get some 'hard' figures rather than having to rely on extrapolation/guesstimates etc so we can get clarification.

 

[BrownTown]

making AMDs scores look worse isn't what they would do since we can test that. Its easier for them to make their scores look better by changing the drivers or game settings to use lower quality settings for their CPU. Nobody will know until mroe people benchmark the CPU on their own instead of on Intel setups.

 

[Sunrise089]

good point Brown, no idea if this is something they would/could easily do, but I agree it seems much more likely.

 

[josh609]

I think you guys are trying to milk a dead cow. I am a big AMD fan, but i have a doubt that those benchmarks are Fake. Anandtech did the benchmarks, not Tomshardware. So i believe it is legit.

 

[TheRyuu]

Originally posted by: josh609
I think you guys are trying to milk a dead cow. I am a big AMD fan, but i have a doubt that those benchmarks are Fake. Anandtech did the benchmarks, not Tomshardware. So i believe it is legit.

Yea, but didn't they do the benchmarks with a rig provided by Intel?

But hey, ya never know. Intel might be that good.

 

[blckgrffn]

Originally posted by: josh609
I think you guys are trying to milk a dead cow. I am a big AMD fan, but i have a doubt that those benchmarks are Fake. Anandtech did the benchmarks, not Tomshardware. So i believe it is legit.

They did it on Intel supplied machines. Once they produce it their own labs, then I will take Anandtech as the truth.

 

[dexvx]

Originally posted by: blckgrffn

Originally posted by: josh609
I think you guys are trying to milk a dead cow. I am a big AMD fan, but i have a doubt that those benchmarks are Fake. Anandtech did the benchmarks, not Tomshardware. So i believe it is legit.

They did it on Intel supplied machines. Once they produce it their own labs, then I will take Anandtech as the truth.

shock
denial <-- AMD fanbois are here
depression
anger
acceptance

 

[plus]

Gee - would this be the first time Intel fudged a benchmark?

 

[Keysplayr]

Originally posted by: dexvx

Originally posted by: blckgrffn

Originally posted by: josh609
I think you guys are trying to milk a dead cow. I am a big AMD fan, but i have a doubt that those benchmarks are Fake. Anandtech did the benchmarks, not Tomshardware. So i believe it is legit.

They did it on Intel supplied machines. Once they produce it their own labs, then I will take Anandtech as the truth.

shock
denial <-- AMD fanbois are here
depression
anger
acceptance

This was funny.

 

[puffpio]

Originally posted by: dexvx
shock
denial <-- AMD fanbois are here
depression
anger
acceptance

shock - "Woah those benchmarks are insane! What if they are priced cheaply?!?!"
denial - "They can't be real, Intel prolly hacked/optimized the demos"
depression - "I just spent all this money on my AMD system and now it's not best price/performance computer..."
anger - "........I just SPENT all this money on my AMD system and now it's not best price/performance computer!!!!!!!!!!"
acceptance - "well, I'll just wait for AMD to overtake Intel again"/"I'm selling my system"/"The hardware industry is always speeding up"


hehehe I'm an AMD fan btw (tho not a fanboi)

 

[Sunrise089]

ehh..I'm an AMD fan, but I don't think any of us are really claiming Conroe is slower than Athlon 64. I just agree that the test shouldn't be taken at 100% face value. Still, my next processor upgrade may very well be to an Intel chip, much as it pains me to say it.

 

[openwheelformula1]

lol image how the Intel fanboys felt over the past years.

 

[sharad]

A company that is several percentages behind and then takes a 40% lead, even in one application, is a little hard to believe.

In any case if the performance numbers are correct and the chip is priced resonably then it is actually good for us. We will have to consider Intel for the next round of upgrade.

 

[openwheelformula1]

It just means significant price cuts for Athlon64, Athlon X2, Pentium 4, and Pentium Ds.

 

[DrMrLordX]

Looks good! Anyone who's been watching Dothan and Yonah benchmarks in the past probably saw this coming. Hell, Dothan had already been shown to be better clock-per-clock in games vs single-core Athlon 64s(by a small margin).

AMD better have something else besides DDR2 support coming. They'll have to get K8 to about the 3.2-3.5 ghz range to beat the 3 ghz EE Conroe I suspect.

Of course, I'd like to see some more benchmarks than those provided . . .

 

[AkumaX]

I am a *insert best price/performance ratio CPU brand HERE* fanboi

 

[boeki]

so, anyone came out with the test yet?

 

[PingSpike]

Well this is going to be an interesting year it seems. I doubt AMD is going to have anything up their sleeves that will knock this out. We'll see what the landscape looks like when these finally come out.

Now the real question...will there be a cheap lowend model I can overclock?

 

[nycdude]

Originally posted by: puffpio

Originally posted by: dexvx
shock
denial <-- AMD fanbois are here
depression
anger
acceptance

shock - "Woah those benchmarks are insane! What if they are priced cheaply?!?!"
denial - "They can't be real, Intel prolly hacked/optimized the demos"
depression - "I just spent all this money on my AMD system and now it's not best price/performance computer..."
anger - "........I just SPENT all this money on my AMD system and now it's not best price/performance computer!!!!!!!!!!"
acceptance - "well, I'll just wait for AMD to overtake Intel again"/"I'm selling my system"/"The hardware industry is always speeding up"


hehehe I'm an AMD fan btw (tho not a fanboi)

Me too. Its been a while since I owned an Intel system.

 

[Zap]

Originally posted by: AkumaX
I am a *insert best price/performance ratio CPU brand HERE* fanboi

Originally posted by: PingSpike
Now the real question...will there be a cheap lowend model I can overclock?

QFT²

How many current AMD fanboys were Intel fanboys back in the early Northwood days of the 1.6A and 1.8A? How many were Intel fanboys back in the early 2.4C days? How many of us now have some kind of "lower end" overclocked socket 754/939 setup? How many of us have multiple computers, some with AMD and some with Intel?

*Zap raises hand*

If there's a cheap overclockable CPU that gives us as good bang for the buck as the slut down the street, then I'm all over it like a sailor on leave.