Original Post
I downloaded the BulletSharp wrapper for BulletPhysics and have been playing with it. I've also run across a problem:
When running DX10 in a test program that doesn't use BulletSharp, it gets ~120 DrawIndexed calls per ms.
When running DX10 in a test program that uses BulletSharp, it gets ~32 DrawIndexed calls per ms.
They're not the same test program, but DX9 runs perfectly fine and much faster with BulletSharp in program #2. So, that's significant.
What could be causing this and how do I fix it? Thanks.
Edit1: I should also note that DX10 is faster in every other test case.
Edit2: Profiling pinpoints the slowdown as happening in the Device.DrawIndexed() and Effect.Apply() functions. Has anyone tested DX10 and BulletSharp/BulletPhysics?
When running DX10 in a test program that doesn't use BulletSharp, it gets ~120 DrawIndexed calls per ms.
When running DX10 in a test program that uses BulletSharp, it gets ~32 DrawIndexed calls per ms.
They're not the same test program, but DX9 runs perfectly fine and much faster with BulletSharp in program #2. So, that's significant.
What could be causing this and how do I fix it? Thanks.
Edit1: I should also note that DX10 is faster in every other test case.
Edit2: Profiling pinpoints the slowdown as happening in the Device.DrawIndexed() and Effect.Apply() functions. Has anyone tested DX10 and BulletSharp/BulletPhysics?
