GPU-MMS

GPU-MMS is a GPU-efficient multiway mergesort implementation. It is the implementation of a research project focused on developing efficient sorting algorithms for modern GPU architectures. This work is published and details are also covered in my PhD dissertation. The implementation of GPU-MMS is in CUDA C++ with the code freely available on github. For more details on this project, please see the published paper, my dissertation, or visit the github page.

Advertisements