A quick PhysX comparison

[memphist0]

A quick PhysX comparison

i7 3770k @ 4.7 GHz, 8 GB @ 1866 MHz cas 11

Batman AC, DX11, FSAA, PhysX High

GTX 580 + GTX 460
Min 30
Max 94
Avg 67
GTX 460 max @ 42% usage

GTX 580 + 8800 GTS
Min 20
Max 95
Avg 55
8800 GTS max @ 71% usage

GTX 580 + GT 640
Min 25
Max 96
Avg 64
GT 640 max @ 59% usage

GTX 580 + GT 620
Min 14
Max 95
Avg 50
GT 620 max @ 79% usage

GTX 580 alone
Min 35
Max 95
Avg 63

GTX 580 alone OC'd
Min 30
Max 96
Avg 65

Borderlands 2, Everything Max'd, PhysX High, Unlimited FPS
Caustic Caverns Recorded with Fraps

GTX 580 + GTX 460
Min 49
Max 98
Avg 69.265
GTX 460 max @ 63% usage

GTX 580 + 8800 GTS
Min 20
Max 79
Avg 36.975
8800 GTS max @ 85% usage

GTX 580 + GT 640
Min 35
Max 92
Avg 68.15
GT 640 max @ 68% usage

GTX 580 + GT 620
Min 16
Max 97
Avg 41.892
GT 620 max @ 88% usage

GTX 580 alone
Min 40
Max 87
Avg 63.942

GTX 580 alone OC'd
Min 44
Max 96
Avg 64.476

I knew the 620 was a weak card (thought it would be nice being single slot) but was surprised it wasn't a better PhysX card, same with the 8800 GTS. Both cards were limiting my 580.

The 460 & 640 added a higher max FPS but average and min stayed the same as the GTX 580 alone. So I figure the playablity would be the same without the 640.

The interesting part about that was before upgrading my CPU from a 2.5 GHz Core 2 I needed the 640 to maintain fps otherwise with the 580 alone fps would drop into the low 30s.

 

[TrailRunner]

This is really cool info, I'm looking forward to seeing those 460 and 8800 numbers too. I'd say that once you get all of the results up that it would be front-page worthy.

 

[DooKey]

You need to get Borderlands 2 and run your tests with that as well. I think you'll see different results.

 

[memphist0]

I ran Borderlands 2 with all settings maxed including PhysX and even with the 580 alone. I was staying right at my fps cap of 72 at a resolution of 1920x1200

 

[memphist0]

Updated with the 460 and 8800 numbers