When comparing GF FX to 9700PRO, why dont benchmarkers do this....

[AgaBoogaBoo]

I think it would *interesting* to see how they fare in the same conditions. OK, you don't have to lower the 9700 down to the FX, instead lower both down a lot. That would show more clearly which cards gets the most done. By doing this, you could also compare which points in each card gave it the most performance boost. You could run tests with a low core clock and a low memory clock. Then one with a high core clock and a low memory clock. Then one with a high memory clock and a low core clock. This can help show what gave each the most performance boost and in future versions, which cards will change their cores faster vs. changing the memory faster.

Its kind of like ripping away the FSB from a P4 down to 33mhz, the P4 would hardly do anything. Whereas, if you ripped away 256k of cache from the 512 the P4 has, it would still fare pretty well, obviously not as well, but the test would prove that the P4 gets the biggest performance boost from a higher FSB. Then you could obviously compare and compare coming up with what you think will be the benchmarks of future processors.

Isn't that interesting?

 

[kylebisme]

exactly, it would be fun and enlighening, if anyone has a gefoce fx to spare i would be happy to hop right to it.

 

[BoomAM]

At Last!!!
We have more that one person that has replied without comparing my suggestion to something.

 

[sash1]

Originally posted by: AgaBooga
Isn't that interesting?

No, becuase what you have mentioned is impossible.

 

[SickBeast]

OK BoomAM, why don't you just find someone on these forums who owns a GFFX and have them benchmark it at 325mhz core speed. Then, if you have a 9700, lower its' memory clock accordingly and run the same benchmark. I know it wouldn't be all that accurate because the two testbeds would be different, but it might satify your curiosity.

 

[bunnyfubbles]

This thread is practically a joke. I have no idea why you seem to require a piece of mind about running the boards at the same "specs." I used to think the same way until I actually learned about the uniqueness of the cores, maybe once you are educated on such matters you will come to understand that running the cards at "equal" specs is still an absurd idea. IF nvidia had engineered the GeForce FX to operate off of insanely high clock speeds (core and ram, and they can reach those speeds with air cooling) then that's how the card was meant to do its job, by attacking video processes with "brute force" (THW). ATI took a different route prefering to use a little more finess to get their speed from bigger bandwidths. ATI did not intend for the 9700 Pro to be operated at insanely high clock speeds. The 9700 Pro would most likely dominate a core crippled GeForce FX as the GPU was meant to be run at 400-500MHz to get comparable performance.

The only thing you could do without lowering to sheer irelevency would be to push the cards as far as they can go in equal environments with equal cooling as it is obvious the GeForce FX has an "unfair" cooling advantage.

Who knows, a Radeon 9700 Pro with 500MHz core 500MHz ram would most likey kick the shiz out of a GeForce FX 500/500, but the fact remains that the 9700 Pro WAS NOT DESIGNED TO OPERATE AT SUCH SPEEDS, the GeForce FX was.

Running them in the same set up with the same cooling solutions and then pushing them as far as they could go while remaining stable would be the only way to test the "equivilency" of the boards.

 

[kylebisme]

bunnyfubbles, i think that should be rephrased to:

Who knows, a Radeon 9700 Pro with 500MHz core 500MHz ram would most likey kick the shiz out of a GeForce FX 500/500, but the fact remains that the 9700 Pro WAS NOT REDESIGNED TO OPERATE AT SUCH SPEEDS, the GeForce FX was.

 

[BoomAM]

Well would you belive it. Yet another person who cant understand the suggestion, and so decides to take out his "confusion rage" on anyone that would find it interesting.

 

[Ketchup]

I think it would be interesting, but people would probably use it to draw conclusions about which card is better, which wouldn't hold true since one of both cards wouldn not be running at the speed at which they are designed to run. It would be like a review I read of an Athlon XP compared to a Pentium 4 of the same speed. Sure the AMD is faster at 2 GHz, but you can't go out and buy a 3 GHz Athlon XP, so the comparison would seem pretty pointless.