Which cards are faster at the same clock rate?

[Munky]

Just out of curiosity, has anyone tried running similar cards at the same clock rate to see which one is faster in a "non-partial" game, like FarCry or CoD (not HL2 or Doom3)? For example, if you downclock a 6600gt to 9800p speeds, which one is faster? Or downclock a x850xt to 6800u speeds. Any one tried this?

 

[2Xtreme21]

Depends on the other features of the card. The 6600gt would theoretically be faster than the 9800pro because of faster RAM (I think) and Shader Model 3.0. It's kinda like asking why an A64 3200+ is faster than an AXP 3200+... still the same clock speed, but as we've found out time and time again with Intel vs. AMD, clock speed isn't even close to everything.

 

[classy]

You can't compare clock rate between video cards or cpus because of different architectures. They are designed differently.

 

[ddogg]

no he knows that..he just trying to see what would happen if lets say u downclock a x850XT from 540 to 400 like a 6800U and then see what would be the results...
good idea, ill try it out since i have both cards!

 

[Keysplayr]

Originally posted by: ddogg
no he knows that..he just trying to see what would happen if lets say u downclock a x850XT from 540 to 400 like a 6800U and then see what would be the results...
good idea, ill try it out since i have both cards!

I have an X800XTPE and a 6800GT. I will downclock the XTPE to 350/1000.
I already know whats going to happen but it's just fun to see. The GT will pretty much landslide the equivalently clocked XTPE. 6xxx series seems to perform better at lower MHz which may indicate a more efficient architecture (similar to how A64's compare to much higher clocked P4's).

ATI fans hate you now that you made this thread. Even if it is just for fun.

EDIT: Yes, I know, i know..... Differents architectures. But that does not change the fact that one architecture has to run at higher clocks to perform as good as another architecture running at slower clocks. This is all good fun guys.

:::::zips up radiation shielded suit. This will go well past SuperNova::::::

 

[McArra]

X850XT @400 core is a X800XL, there you have your answer

 

[Munky]

Well, see it could also be like the p4 vs a64, where one was designed for high ipc, and the other designed for high mhz. So, if Nvidia's archtecture is faster at the same clock rate, that would also mean it would be harder to reach the higher mhz of Ati's cards. Not that one is necessarily better than the other, there's a lot of other factors involved, but it would be an interesting test.

I'm curious, though, because a 6600 vanilla is slower than a 9800p, but the 6600gt is faster. So, given that the only difference between the 6600 series is clock rate, how would a 6600gt compare to a 9800p is you downclocked it to the same level. Also, what would hurt it more, downclocking the gpu or the memory?

Edit: actually since the 6600gt is a 128bit memory bus, you can't really compare them fairly at the same mem speed. But if a x850 @400 = x800xl, I'm guessing Nvidia will be slightly faster at 400.

 

[ddogg]

Originally posted by: keysplayr2003

Originally posted by: ddogg
no he knows that..he just trying to see what would happen if lets say u downclock a x850XT from 540 to 400 like a 6800U and then see what would be the results...
good idea, ill try it out since i have both cards!

I have an X800XTPE and a 6800GT. I will downclock the XTPE to 350/1000.
I already know whats going to happen but it's just fun to see. The GT will pretty much landslide the equivalently clocked XTPE. 6xxx series seems to perform better at lower MHz which may indicate a more efficient architecture (similar to how A64's compare to much higher clocked P4's).

ATI fans hate you now that you made this thread. Even if it is just for fun.

EDIT: Yes, I know, i know..... Differents architectures. But that does not change the fact that one architecture has to run at higher clocks to perform as good as another architecture running at slower clocks. This is all good fun guys.

:::::zips up radiation shielded suit. This will go well past SuperNova::::::

well it definetely means a much more efficient architecture...geez the X850XTPE has a 140MHZ advantage yet it barely outperforms the 6800U. even though they are different architectures u cant rule out such a big difference!!

EDIT: anyone got a X800pro and a 6800GT??...wonder what would happen running the Pro at GT speeds just to see the performance difference

 

[Keysplayr]

Originally posted by: ddogg

Originally posted by: keysplayr2003

Originally posted by: ddogg
no he knows that..he just trying to see what would happen if lets say u downclock a x850XT from 540 to 400 like a 6800U and then see what would be the results...
good idea, ill try it out since i have both cards!

I have an X800XTPE and a 6800GT. I will downclock the XTPE to 350/1000.
I already know whats going to happen but it's just fun to see. The GT will pretty much landslide the equivalently clocked XTPE. 6xxx series seems to perform better at lower MHz which may indicate a more efficient architecture (similar to how A64's compare to much higher clocked P4's).

ATI fans hate you now that you made this thread. Even if it is just for fun.

EDIT: Yes, I know, i know..... Differents architectures. But that does not change the fact that one architecture has to run at higher clocks to perform as good as another architecture running at slower clocks. This is all good fun guys.

:::::zips up radiation shielded suit. This will go well past SuperNova::::::

well it definetely means a much more efficient architecture...geez the X850XTPE has a 140MHZ advantage yet it barely outperforms the 6800U. even though they are different architectures u cant rule out such a big difference!!

EDIT: anyone got a X800pro and a 6800GT??...wonder what would happen running the Pro at GT speeds just to see the performance difference

No wait!!! An X800pro clocked the same as a 6800nu with 256MB. both
are 12 pipe/256bit/256MB

Clock the X800pro at 325/600. I would love to see that.

 

[imported_Noob]

Originally posted by: munky
Just out of curiosity, has anyone tried running similar cards at the same clock rate to see which one is faster in a "non-partial" game, like FarCry or CoD (not HL2 or Doom3)? For example, if you downclock a 6600gt to 9800p speeds, which one is faster? Or downclock a x850xt to 6800u speeds. Any one tried this?

Well I would think the 6800U would perform faster if you set the GPU clock on an X850 XT to 400MHz.

 

[BillyBobJoel71]

yes it would becasue the clock speeds on ati cards have always been higher thatn there nvidia equal competitors, yet they perform the same or worse. its the different gpu architecture of the chips.

 

[BFG10K]

NV4x would be faster compared to the R4xx at the same core clock.

 

[Gamingphreek]

Originally posted by: BFG10K
NV4x would be faster compared to the R4xx at the same core clock.

Yes, as for comparing 6600's to 9800's that is where the problem occurs. You are comparing different generations of cards. The 9800Pro's are 256bit data width while the 6600 is 128. To do this you mist keep all other variables the same, (ie if you change the clockspeed, you have to compare with the same number of pipes).

I agree with BFG.

-Kevin