General FPS-benchmark question

Forum for all the nVidia video cards from the past, present and future!
Post Reply
R3N3G4D3
Legit Fanatic
Legit Fanatic
Posts: 121
Joined: Fri Dec 23, 2005 2:06 pm

General FPS-benchmark question

Post by R3N3G4D3 »

Hey guys,

I've always been referring to legit reviews for every new graphics card that comes out when I would consider purchasing. Now I've bought a new 8800GTS 512 (after my old one died :( ) and this baby seems to be overclocking much better (now running 824/1103 stable on stock), although I can't push the memory higher so it seems to still be Qimonda. Anyway, I'm planning to volt-mod it (because one dead 8800GTS 512 is not enough :roll: ) and I wanted to benchmark-compare it to the stock 8800GTS 512.

The problem is I'm not sure how to obtain consistent results, so I figured I'd ask from the guys who're used to doing graphics card reviews. Currently I setup RivaTuner to catch ingame FPS and output them on the screen during the game. But it only outputs the current FPS, which fluctuates a lot. How can I get the min/avg/max fps for the game without sitting there and constantly writing them down with a pencil. I also don't trust RivaTuner's logging since it logs everything including the in-menu FPS which are way off from the game ones (my UT3 runs at 125-175 ingame, and in menus can show anything from 0 to 700 fps).
User avatar
Apoptosis
Site Admin
Site Admin
Posts: 33941
Joined: Sun Oct 05, 2003 8:45 pm
Location: St. Louis, Missouri
Contact:

Re: General FPS-benchmark question

Post by Apoptosis »

When we test video cards we test them on a clean test system with the latest patches and game updates. We then start up the system and wait for Vista 64 to cache all available memory... I. This takes between 5-7 minutes after you log on to your account, you might also want to disable drive indexing on the C: drive. f you are connected to the internet you might want to disable the connection or pull the cable, unless you are testing games like Lost Planet that need internet We then run the time demo, benchmark script or manually run fraps after this point on one game. Run the benchmark 3 times, restart the whole process over two more times. At the end you'll have nine benchmark runs to average. You'll see consistent numbers like this.
R3N3G4D3
Legit Fanatic
Legit Fanatic
Posts: 121
Joined: Fri Dec 23, 2005 2:06 pm

Re: General FPS-benchmark question

Post by R3N3G4D3 »

Thanks Apoptosis,

I'm probably not going to reinstall the OS since this is just for my own benchmark instead of official one, but aside from that I'll try to follow all of the steps. Fraps does seem to have much better logging features than RivaTuner, so I'll use that instead.
User avatar
vicaphit
Legit Extremist
Legit Extremist
Posts: 2789
Joined: Thu Apr 05, 2007 2:22 pm
Location: SE Ohio

Re: General FPS-benchmark question

Post by vicaphit »

I have seen some performance decreases with fraps running though.
----
Andy

New Rig: Intel Q6600 (2.7ghz), ASUS P5KC, 4GB DDR2-800, Palit 9600GT (1GB), Ultra X3 1KW (thanks LR!), Vista 64bit
Dell Rig: AMD Athlon X2 4000+ (2.1ghz), 2 GB DDR2 Ram, MSI 8800GT (512MB), Vista 32bit

PSN: VICaphit
Post Reply