A Comparison of the performance of HPC Accelerators

This project aims to port the scientific application GADGET-3 to multiple accelerators, research on the performance achieved and compare the porting/optimisations on the given accelerators with different architectures. In this project, the most time-consuming functions of GADGET-3 was identified based on the profiling. Partial functions in GADGET-3 were ported to the accelerator NVIDIA K40 card using CUDA C and to AMD FirePro W9100 Card using OpenCL. The CUDA version GADGET-3 functions were profiled and analysed with different task assignment on the GPGPU device for the given test cases. Detailed comparisons were also conducted between the CUDA C version GADGET-3 on NVIDA cards and the OpenCL version code on AMD card. For the given test cases, NVIDAI K40 delivered better performance compared with AMD FirePro W9100.