Branch Predication in CUDA Any example in SDK

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi,

I want to understand the concept of Branch Predications in CUDA. Is there any example source code in SDK or elsewhere I could learn it ? Even any introductory explanation will suffice for my case.

Thanks

2

I don’t think that predication is exposed at the CUDA level; ideally, the compiler should do it for you. You would have to write in PTX to use it right now I think…