Forcing serail computation

Accelerated Computing CUDA CUDA Programming and Performance
1

Does anybody know of a quick way, for debugging purposes, to force the GPU into performing all computations serially?
i.e., to invoke only one thread at a time?

Thanks,
Jon.

2

Try to run on device emulation mode with cuda 3.0 or early. It has warp size 1. It may help. Also compare debug and release mode too.

3

Try to run on device emulation mode with cuda 3.0 or early. It has warp size 1. It may help. Also compare debug and release mode too.

4

Thanks… Didn’t think of that :)

5

Thanks… Didn’t think of that :)