View Single Post
Old 05-15-03, 08:01 AM   #110
Registered User
Join Date: Aug 2002
Posts: 114

Originally posted by nVidi0t No, because I have tested it, and major differences to system spec, but not changing the display card does not affect performance in the benchmark, yet in games, the average FPS is 50% faster.
That's because it's a VID-CARD-test, not CPU-test! With previous 3DMArk, they were criticized that the benchmark is being held back by the CPU. That's why the new 3DMark is alot more vid-card dependant.

It shows all the specs, and allows you to compare your results to other users with the same CPU, obviously it's a system comparison.
Since it does have tests that are not vid-card dependant (like the sound-test), it's a good thing to show all relevant specs. Besides, showing the specs does no harm, so why not do it?

The sound tests do impact score.
WRONG! Here's what Futuremark says about this (page 20 of the whitepaper):

"Each of the four game tests generates a frame-rate that is used to calculate the overall score"

It even show the formula that they usaed to calculate the score!

And the CPU tests should aswell, yet trying an exact system spec with another CPU almost 50% faster, the 3dmark scores remain the same.
As it should be since it's a VID-CARD BENCHMARK!

It's a bad benchmark. Game engine benchmarks are the only real way of determining performance WHICH MATTERS to the customer, they are going to be playing games, not benchmarks. (unless they are like a few peeps here.. lol)
NV didn't have any problems with previous 3DMarks when they were perfectly suited for their feature-set. Now that that is not the case, they whine that "only game-benchmarks matter". hypocrisy, plain and simple.
Nemesis77 is offline