关键词: LAREDP, Intel Xeon Phi, Native模式, Offload模式, 512位SIMD扩展指令, 异步数据传输, 双缓冲

Abstract:

Plasma simulations have been widely used to exploit scientific problems under extremely situations. The paper ports a particleincell based plasma code, LAREDP, to Intel Xeon Phi coprocessor. In order to accomplish this, two modes are employed, i.e. the Native mode and the Offload mode. Firstly, the Native mode is employed to study on the hot computing tasks, which have been accelerated to 4.61 times faster by using SIMD extension instructions. Secondly, the Offload mode is employed to transplant the whole code onto a CPUIntel Xeon Phi heterogeneous system. We also adopt optimizations such as asynchronous data transferring and double buffer technique to improve the performance. And we obtain 9.8% and 21.8% improvement respectively.
    

Key words: LARED-P;Intel Xeon Phi;native mode;offload mode;512bit SIMD extension;asynchronous data transferring;double buffer