Is it possible to synchronize CUDA streams to perform a multipart computation?

Question

I\'m trying to simulate a spiking neural network, which requires two separate kernels, 1 kernel to run the sensory neurons, and a 2nd kernel to run the interneurons. I\'m trying