Capturing cuda events in a kernel

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi, is there a way to capture occurrence of a CUDA event from within a kernel?

I’d like to have something like this:

__global__ void kernelK(...) {

while(..) {

      cudaEventSynchronize(eventXYZ);

      //doSomeWork 

      ...

    }

}

Is this possible on current GPUs, and what should I take care about?

2

Currently there is no high-level structure that covers this for you. You’d have to implement the desired behaviour on your own.