I can't find a really straight answer on this via Google, but I was wondering if there are cases where you get better performance by disabling Nvidia's PhysX function on the GPU and allowing the CPU to deal with the physics. For example, you might have a lower-end or older NVidia card that technically "supports" PhysX, and a high-end multicore CPU. Anyone play around with this? I'm talking about a single-GPU situation, not SLI. Thanks for any thoughts.
My Computer
- OS
- Win7 Home Premium x64
- CPU
- Intel Core i7/860
- Motherboard
- Gigabyte GA-P55M-UD2
- Memory
- 4 GB PC 1333
- Graphics Card(s)
- Nvidia GTS 250 (1 GB mem)
- Sound Card
- Onboard Realtek ALC888B
- Hard Drives
- Western Digital 500 gig, 7200 rpm
- PSU
- 700W