Advanced HIP Workshop | Optimising compute with concurrent IO | Sequential (synchronous) IO solution

This video is part of the HIP workshop playlist. View full playlist here: • Advanced HIP Workshop
We use an array of three HIP device allocations to represent the wavefield at various timesteps. Notice that with the hipMemcpy call above, we used an address located within the host memory allocation out_h to perform the copy. This technique is not valid for use in asynchronous copies.