Benchmarking programs

[Scared0o0Rabbit]

I wanna take some benchmarks before and after this graphics card upgrade. I've used 3dmark2001 in the past, however I'm sure there are much more current ones out there, and would like suggestions as to one to try. I'd prefer if said program was free.

Edit: there seems to be a 3dmark2003, I'm downloading it right now, here's to hoping my current video card is good enough to actually run it lol.

 

[Curtis]

Aquamark is another one to consider. It should run on most g/c hardware in one form or another.

3Dmark03 should run, but don't expect much from it. Using my old GF2MX it only ran the first test + the fillrate test + the sound tests and returned a score of ~150. After an upgrade to a 9800Pro, it jumped to ~5400.

 

[Scared0o0Rabbit]

Yeah I had it do the smae thing to my radeon 7200... with my very lsight overclock on the radeon I got a score of 189 lol. I imagine it'll jump up when I get my new card.

 

[gameboy900]

3DMark2K1, 3DMark2K3 and AquaMark are the current benchmarking tools for vid cards. You can also use their comparison features to see how your systems stacks up against others with similar specs.

Also if you're gonna run the tests the best method to get a reliable score is to run each one 3 times in a row then take the average of the three scores. This helps to compensate for any slight deviations that may have occured in one of the tests.

For the two 3DMarks you ONLY need to do the Game tests to get a score. The other tests do not affect the final number. Just something to speed up benchmarking for you.

 

[Alexvrb]

Remember that canned benchmarks like those often don't correspond to real world performance with actual games. I would rather run real games that have benchmark capabilities built in. You didn't happen to be under a rock when Nvidia and Intel were shelling out money to "give input" to Futuremark? They aren't paying money to be on their board for no reason.

 

[Tagrineth]

Originally posted by Alexvrb@Jan 6, 2004 @ 12:23 AM

Remember that canned benchmarks like those often don't correspond to real world performance with actual games. I would rather run real games that have benchmark capabilities built in. You didn't happen to be under a rock when Nvidia and Intel were shelling out money to "give input" to Futuremark? They aren't paying money to be on their board for no reason.

lol... actually nVidia was more famous for bitching about how expensive that was, and thus NOT being on Futuremark's beta crew.

Hence the "REAL REASON" (lol!) for why the GeForce FX series is so pathetically slow in 3DMark03 when the plethora of cheats gets bypassed...

 

[Alexvrb]

They kissed and made up. In a very obvious way... shortly after Futuremark's initial anti-Nvidia-cheating press releases, they had another release talking about acceptable optimizations that don't affect EQ, etc. So I really don't trust synthetic benches for comparing hardware from different vendors, or even really different designs. It's such a mess.