What could cause a change in fps?

Kelv · Jul 10, 2008

I don't know what I was doing on June 30th, but I eventually did something good. This is the Crysis Benchmark. I was overclocking one of the 4850's and I eventually stuck with 660-1050, and these are the results that I got. I marked these both as a single 660-1050 oc. I'm 90% sure the second result was not with crossfire. These were with the hotfix drivers.

Code:

6/30/2008 7:54:17 PM - XP
Beginning Run #1 on Map-island, Demo-benchmark_gpu
DX9 1280x1024, AA=4x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: High
 ==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
    Play Time: 56.37s, Average FPS: 35.48
    Min FPS: 23.01 at frame 138, Max FPS: 46.11 at frame 991
    Average Tri/Sec: 36015772, Tri/Frame: 1015026
    Recorded/Played Tris ratio: 0.90
!TimeDemo Run 1 Finished.
    Play Time: 51.52s, Average FPS: 38.82
    Min FPS: 23.01 at frame 138, Max FPS: 46.66 at frame 1014
    Average Tri/Sec: 39846132, Tri/Frame: 1026366
    Recorded/Played Tris ratio: 0.89
!TimeDemo Run 2 Finished.
    Play Time: 51.48s, Average FPS: 38.85
    Min FPS: 23.01 at frame 138, Max FPS: 46.89 at frame 1012
    Average Tri/Sec: 39874520, Tri/Frame: 1026271
    Recorded/Played Tris ratio: 0.89
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Completed All Tests

<><><><><><><><><><><><><>>--SUMMARY--<<><><><><><><><><><><><><>

6/30/2008 7:54:17 PM - XP

Run #1- DX9 1280x1024 AA=4x, 32 bit test, Quality: High ~~ Overall Average FPS: 38.835

This was just an hour later. I really wish I could remember what I was doing but I almost think that I was just playing Crysis and then decided to run the benchmark again and see how it was doing.

Code:

6/30/2008 8:59:58 PM - XP
Beginning Run #1 on Map-island, Demo-benchmark_gpu
DX9 1280x1024, AA=4x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: High
 ==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
    Play Time: 53.82s, Average FPS: 37.16
    Min FPS: 21.67 at frame 146, Max FPS: 49.25 at frame 993
    Average Tri/Sec: -32983786, Tri/Frame: -887630
    Recorded/Played Tris ratio: -1.03
!TimeDemo Run 1 Finished.
    Play Time: 48.74s, Average FPS: 41.03
    Min FPS: 21.67 at frame 146, Max FPS: 49.77 at frame 984
    Average Tri/Sec: -36375704, Tri/Frame: -886521
    Recorded/Played Tris ratio: -1.03
!TimeDemo Run 2 Finished.
    Play Time: 48.62s, Average FPS: 41.13
    Min FPS: 21.67 at frame 146, Max FPS: 49.77 at frame 984
    Average Tri/Sec: -36460596, Tri/Frame: -886439
    Recorded/Played Tris ratio: -1.03
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Completed All Tests

<><><><><><><><><><><><><>>--SUMMARY--<<><><><><><><><><><><><><>

6/30/2008 8:59:58 PM - XP

Run #1- DX9 1280x1024 AA=4x, 32 bit test, Quality: High ~~ Overall Average FPS: 41.08

Heres the card with stock settings.

Code:

6/30/2008 2:37:21 PM - XP
Beginning Run #1 on Map-island, Demo-benchmark_gpu
DX9 1280x1024, AA=4x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: High
 ==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
    Play Time: 62.44s, Average FPS: 32.03
    Min FPS: 24.55 at frame 162, Max FPS: 41.90 at frame 1002
    Average Tri/Sec: 32530860, Tri/Frame: 1015598
    Recorded/Played Tris ratio: 0.90
!TimeDemo Run 1 Finished.
    Play Time: 55.11s, Average FPS: 36.29
    Min FPS: 24.55 at frame 162, Max FPS: 43.61 at frame 1007
    Average Tri/Sec: 37259296, Tri/Frame: 1026626
    Recorded/Played Tris ratio: 0.89
!TimeDemo Run 2 Finished.
    Play Time: 55.14s, Average FPS: 36.27
    Min FPS: 24.55 at frame 162, Max FPS: 43.61 at frame 1007
    Average Tri/Sec: 37261708, Tri/Frame: 1027313
    Recorded/Played Tris ratio: 0.89
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Completed All Tests

<><><><><><><><><><><><><>>--SUMMARY--<<><><><><><><><><><><><><>

6/30/2008 2:37:21 PM - XP

Run #1- DX9 1280x1024 AA=4x, 32 bit test, Quality: High ~~ Overall Average FPS: 36.28

The only reason why I am posting this is because I can't replicate the fps that I made on June 30th.

Besides for my main question, I was also wondering if anyone could test the fps that they get in this benchmark with the same oc. Thanks for reading.

Update: I'm beginning to lead to the idea that maybe I screwed up and didn't know that I had crossfire enabled. The reason why is because my crossfire results almost mimic the results of the test on June 30th. Only slightly better. Back then, I thought that crossfire slightly hurt the performance in Crysis. I know I sound confused but I ran several tests in which the framerate was a couple 10tenths of a fps less. However, I am receiving a nice bonus from crossfire atm. I recently tried the 8.7 drivers, and I noticed the nice crossfire bonus in that. But when I tested a single card, I was only getting 38 fps which was less than the time I thought I got 41. So I rolled back the drivers and decided to run the driver cleaner along with only installing the hotfix instead of installing the 8.6 and then updating to the hotfix.

The one number that leads me to believe that it must of been crossfire is the Recorded/Played Tris ratio: -1.03. What's the chance that those numbers could be the same?

Code:

7/10/2008 3:20:31 PM - XP
Beginning Run #1 on Map-island, Demo-benchmark_gpu
DX9 1280x1024, AA=4x, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: High
 ==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
    Play Time: 52.91s, Average FPS: 37.80
    Min FPS: 22.72 at frame 149, Max FPS: 49.86 at frame 1011
    Average Tri/Sec: -33574292, Tri/Frame: -888227
    Recorded/Played Tris ratio: -1.03
!TimeDemo Run 1 Finished.
    Play Time: 48.11s, Average FPS: 41.57
    Min FPS: 22.72 at frame 149, Max FPS: 51.30 at frame 986
    Average Tri/Sec: -36850000, Tri/Frame: -886462
    Recorded/Played Tris ratio: -1.03
!TimeDemo Run 2 Finished.
    Play Time: 47.75s, Average FPS: 41.88
    Min FPS: 22.72 at frame 149, Max FPS: 51.30 at frame 986
    Average Tri/Sec: -37097804, Tri/Frame: -885726
    Recorded/Played Tris ratio: -1.03
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Completed All Tests

<><><><><><><><><><><><><>>--SUMMARY--<<><><><><><><><><><><><><>

7/10/2008 3:20:31 PM - XP

Run #1- DX9 1280x1024 AA=4x, 32 bit test, Quality: High ~~ Overall Average FPS: 41.725

Krenum · Jul 10, 2008

Dont read too much into Benchmarking Crysis man, its an unoptimized game, so benchmarking it is pretty much moot.

vgkarthik88 · Jul 10, 2008

crysis is the 1 game that you will never see good scaling with. itsjust like that. try some other game to verify your OCing. also, drivers dont scale as well on Xp as they do on vista. that is another thing u have to keep in mind. i hope ur not using the Catalyst 8.7 leaked beta! haha man those are prob some crap drivers. jus stick to 8.6 + hotfix. those drivers could be for AGP cards or something ! lolz!

What could cause a change in fps?

Kelv

Limp Gawd

Krenum

Fully [H]

vgkarthik88

n00b