Initial results on computational performance of Intel many integrated core, sandy bridge, and graphical processing unit architectures: implementation of a 1D c++/OpenMP electrostatic particle-in-cell code

摘要

We present initial comparison performance results for Intel many integrated core (MIC), Sandy Bridge (SB),rnand graphical processing unit (GPU). A 1D explicit electrostatic particle-in-cell code is used to simulate arntwo-stream instability in plasma. We compare the computation times for various number of cores/threadsrnand compiler options. The parallelization is implemented via OpenMP with a maximum thread number ofrn128. Parallelization and vectorization on the GPU is achieved with modifying the code syntax for compatibilityrnwith CUDA. We assess the speedup due to various auto-vectorization and optimization level compilerrnoptions. Our results show that the MIC is several times slower than SB for a single thread, and it becomesrnfaster than SB when the number of cores increases with vectorization switched on. The compute times forrnthe GPU are consistently about six to seven times faster than the ones for MIC. Compared with SB, thernGPU is about two times faster for a single thread and about an order of magnitude faster for 128 threads.rnThe net speedup, however, for MIC and GPU are almost the same. An initial attempt to offload parts of therncode to the MIC coprocessor shows that there is an optimal number of threads where the speedup reaches arnmaximum.