A Mirror Edge based benchmark was done about 2 months ago... it showed that the original PhysX card was better than high-end video cards. That higher end video cards do much better than entry ones.
However, I personally questioned some of the numbers... I would like to see more independent tests to confirm. I would also like to see someone test the impact of number of SP, core speeds, memory size/speed, ect on PhysX.
However, I personally questioned some of the numbers... I would like to see more independent tests to confirm. I would also like to see someone test the impact of number of SP, core speeds, memory size/speed, ect on PhysX.