Cooperative Groups: Flexible CUDA Thread Programming

Published: 5/10/2017
Cooperative Groups: Flexible CUDA Thread Programming
Source: DEVBLOGS.NVIDIA.COM

In efficient parallel algorithms, threads cooperate and share data to perform collective computations. To share data, the threads must synchronize. The granularity of sharing varies from algorithm to algorithm, so thread synchronization should be flexible. Making synchronization an explicit part of the program ensures safety, maintainability, and modularity. CUDA 9 introduces Cooperative Groups, w

Read more
Related news
Comment