Nvidia physx driver shouldn't matter, but I included that factor in the title anyways
I just installed and did a few benchmarks of my current rig in crysis gpu benchmark, while I know it is far from the be-all and the end-all of benches, I think I saw some pretty big increases in fps in this forceware update, let me know if these numbers seem consistant with previous numbers or if this does indeed seem a lot higher
directx10, high spec, no AA, default nvidia global control panel settings (default AF), 1680*1050
This is from the 177.39 forceware today, taken from the gpu benchmark log:
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 51.38s, Average FPS: 38.92
Min FPS: 20.29 at frame 1937, Max FPS: 52.70 at frame 945
Average Tri/Sec: -26421664, Tri/Frame: -678825
Recorded/Played Tris ratio: -1.35
!TimeDemo Run 1 Finished.
Play Time: 45.49s, Average FPS: 43.96
Min FPS: 20.29 at frame 1937, Max FPS: 53.80 at frame 908
Average Tri/Sec: -29791576, Tri/Frame: -677624
Recorded/Played Tris ratio: -1.35
I did just 2 timedemos, I noticed a bit of microstutter in the first timedemo which is likely why the fps is a little lower
This isn't a epeen glory thread, I was just wondering if this was an increase or if it seemed like perfectly consistant with the previous forceware
1680*1050, 0xAA, 16xAF, High
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 50.98s, Average FPS: 39.23
Min FPS: 20.38 at frame 1919, Max FPS: 53.93 at frame 897
Average Tri/Sec: -26703790, Tri/Frame: -680645
Recorded/Played Tris ratio: -1.35
!TimeDemo Run 1 Finished.
Play Time: 46.01s, Average FPS: 43.46
Min FPS: 20.38 at frame 1919, Max FPS: 56.47 at frame 79
Average Tri/Sec: -29430302, Tri/Frame: -677103
Recorded/Played Tris ratio: -1.35
1680*1050, 0xAA, 16xAF, very high
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 73.62s, Average FPS: 27.17
Min FPS: 16.45 at frame 1928, Max FPS: 35.91 at frame 898
Average Tri/Sec: -9433012, Tri/Frame: -347219
Recorded/Played Tris ratio: -2.64
!TimeDemo Run 1 Finished.
Play Time: 73.44s, Average FPS: 27.23
Min FPS: 15.96 at frame 803, Max FPS: 35.91 at frame 898
Average Tri/Sec: -9039368, Tri/Frame: -331910
Recorded/Played Tris ratio: -2.76
Message edited by ovaltineplease on 06-20-2008 at 11:52:27 PM
1280*1024, Very high spec, 4x AA, 16xAF - same gpu benchmark
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 85.33s, Average FPS: 23.44
Min FPS: 9.63 at frame 1923, Max FPS: 29.94 at frame 382
Average Tri/Sec: -13868222, Tri/Frame: -591720
Recorded/Played Tris ratio: -1.55
!TimeDemo Run 1 Finished.
Play Time: 80.91s, Average FPS: 24.72
Min FPS: 9.63 at frame 1923, Max FPS: 32.25 at frame 88
Average Tri/Sec: -14302641, Tri/Frame: -578647
Recorded/Played Tris ratio: -1.58
1280*1024, Very high spec, 0xAA, 16xAF
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 56.32s, Average FPS: 35.51
Min FPS: 20.33 at frame 1975, Max FPS: 47.59 at frame 880
Average Tri/Sec: -21664610, Tri/Frame: -610073
Recorded/Played Tris ratio: -1.50
!TimeDemo Run 1 Finished.
Play Time: 51.05s, Average FPS: 39.18
Min FPS: 16.86 at frame 1055, Max FPS: 47.73 at frame 997
Average Tri/Sec: -23432092, Tri/Frame: -598121
Recorded/Played Tris ratio: -1.53
Message edited by ovaltineplease on 06-20-2008 at 11:39:51 PM
""Nvidia promised that it will deliver PhysX API compatible with all GeForce 8 and 9 cards in July. For now, PhysX only works on the GeForce models GTX 260, 280, 9800GT, 9800GTX, 9800GTX+ and 9800GX2.""
yea, I don't think the physx is working - at least not in the ut3 physx ctf maps - it just hangs
The driver, uh its through a hardocp video->nvidia forum link but you might have to do a lot of hunting to get one that isn't corrupted because of the website loads - suffice to say I got lucky with hardocp's link.
I /think/ I am seeing an increase in performance but i'm not 100% positive because I haven't erased my gpu_benchmark log in ages so its hard to say what some of the older results' settings were unless I uninstall drivers->reinstall->do more benchmarks
Other than Unreal Tournament 3, there aren't any major titles that use Ageia physix. Game developers perfer their own generic physics engine that's optimized to run on cpu instead. As for your Crysis demo, Crysis, like nearly everyone else, run their physics engine off cpu.
Other than Unreal Tournament 3, there aren't any major titles that use Ageia physix. Game developers perfer their own generic physics engine that's optimized to run on cpu instead. As for your Crysis demo, Crysis, like nearly everyone else, run their physics engine off cpu.
yea, I know that, I listed the physx driver as being installed because its part of the system configuration regardless
While i'm not going to bother listing every program on the pc, that is still pertinent information. I understand what you mean, but its just thrown on there because it was released at the same time as this driver and having the physx installed could either a: inhibit performance (makes the UT3 physx maps freeze on load), or b: do nothing, (doesn't effect crysis)
I think physx will die out eventually. Havok will be more widely supported.
Hmm, hard to say - I could see Epic licensing it as an addon to their engine for UE4, but that is just speculating.
Quote :
What are you talkina bout? Mass Effect, Gears Of War, Age of empires 3, UT3 + UT3 physics mod
All awesome games
I own ME, GoW, UT3 - I like ME/UT3 but after playing through GoW once on hardcore I dropped it - although I hear multiplayer was what made GoW good, not SP so much - I found GoW SP frustrating at times on hardcore, but I got through it.
I didn't know though that GoW and Mass Effect had Physx support - how does it work?
heh, yea I figured Nvidia was just blowing smoke with that - I checked that page before and I was like "yea thats cute except any of those games i've played don't make big use of physics.."
trust me I took that list with a big block of salt lol
I guess maybe if you were running a lower end PC and playing those games then the physx card might make a bigger difference? I wouldn't really know though.
i have an nf4 a8n32-sli-deluxe mb and i currently have one 8800 gts 512 (g92) with another on the way in the mail.
when it gets here i will of course be running in SLI.
my question is....what about buying the asus ageia card and putting it in my pci-e 4x slot....the card will work in even though its a pci-e 1x card.
i want to know if i forced the programs to use ageia instead of nvidia physx if it would bottleneck my system AND would i see a pretty good frame rate increase by having the ageia card dedicated to all the physx processing instead of doing the physx on my nvidia 8800 gts cards.
i have an nf4 a8n32-sli-deluxe mb and i currently have one 8800 gts 512 (g92) with another on the way in the mail.
when it gets here i will of course be running in SLI.
my question is....what about buying the asus ageia card and putting it in my pci-e 4x slot....the card will work in even though its a pci-e 1x card.
i want to know if i forced the programs to use ageia instead of nvidia physx if it would bottleneck my system AND would i see a pretty good frame rate increase by having the ageia card dedicated to all the physx processing instead of doing the physx on my nvidia 8800 gts cards.
thanks.
You PCIE 1x will work on a 4x slot. You don't need to force ageia to work, just select it in the drivers.
------------------------------Rock journalism is people who can't write interviewing people who can't talk for people who can't read - Frank Zappa
Reply to radnor
""Nvidia promised that it will deliver PhysX API compatible with all GeForce 8 and 9 cards in July. For now, PhysX only works on the GeForce models GTX 260, 280, 9800GT, 9800GTX, 9800GTX+ and 9800GX2.""
You are about to answer a thread that has been inactive for more than 6 months. If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.