Break down GPU execution time with an analytical method

Junjie Lai, André Seznec · 2012

Because modern GPGPU can provide significant computing power and has very high memory bandwidth, and also, developer-friendly programming interfaces such as CUDA have been introduced, GPGPU becomes more and more accepted in the HPC research area. Much research has been done to help developers to better optimize GPU applications. But to fully understand GPU performance behavior remains a hot research topic.

Read the paper · More papers on PaperTik