PhysX On A Dedicated GPU

The final feature to be introduced in the Release 180 ForceWare drivers is the ability to assign a dedicated graphics card for PhysX acceleration. This is pretty cool since it allows you to reutilize an older graphics card which would normally be lying dormant.

Previously, you can only choose to use a single graphics card or multiple cards of the same model in SLI mode. PhysX calculations would be automatically processed together with 3D graphics. With the R180 drivers, you can now add a separate card (of a different model) to process PhysX calculations. This allows the primary graphics card to concentate on rendering 3D graphics.

NVIDIA says that any NVIDIA DirectX 10 graphics card (GeForce 8 onwards) can be used for dedicated PhysX acceleration. However, it should have at least 256 MB of graphics memory. Using this feature will give old NVIDIA graphics cards new life. It also frees up primary graphics card to do pure 3D graphics rendering.

NVIDIA's results in Warmonger showed that using a GeForce 9600 GT to perform dedicated PhysX rendering can provide a significant (28%-42%) boost in frame rates over just using a single graphics card to both render graphics and process PhysX calculations. Of course, we have no idea under what conditions NVIDIA obtained those results.

Support Tech ARP!

If you like our work, you can help support out work by visiting our sponsors, participate in the Tech ARP Forums, or even donate to our fund. Any help you can render is greatly appreciated!