|国家预印本平台
首页|基于国产众核处理器三维地震声波正演模拟

基于国产众核处理器三维地震声波正演模拟

中文摘要英文摘要

三维地震声波理论与计算方法是地质勘探研究的基础,通过分析不同介质中声波的传播特性,完成三维地震声波正演模拟。针对三维地震声波有限差分交错网格方程正演过中存在数值计算大、内存消耗大等实际问题,本文提出基于“神威·太湖之光”超级计算机系统中,国产异构众核处理器(申威26010)的三维地震声波正演模拟编程模型,完成基于处理器间的进程级并行,与基于计算核心间的线程级并行优化策略。研究了DMA(直接内存读取)通信方式,提出2.5D流水线任务划分、通信与计算的相互掩盖的多角度的优化策略。实验结果表明该策略有效缓解了带宽瓶颈,发挥了处理器强大的计算能力,解决了程序在申威26010异构众核处理器处理有限差分问题时,并行效率过低的问题。在大规模测试下,使用266240个计算核心,程序仍能够保持稳定的计算性能,达到5.5GFlops的场值更新。

3d seismic acoustic wave theory and calculation method are the basis of geological exploration research. By analyzing the propagation characteristics of acoustic waves in different media, we can apply the 3d seismic acoustic wave forward mod-eling in exploration work. In order to solve the problem of huge numerical calculation and large memory consumption while proceeding 3D seismic wave equation staggered grid finite difference forwarding model, we studied and implemented the parallel optimization on the heterogeneous many-core processors of the Sunway Taihulight supercomputer. Based on the implementation of a two-level parallel programming model by using MPI+Sunway Athread, we generated the DMA communication, 2.5D pipelining task division and other optimization strategies. The model with this improvement reduced the negative effects by bandwidth and greatly utilized the computing power. In large-scale conditions, we tackled the issue of low efficiency about program execution on SW26010 het-erogeneous many-core processors. The experimental results reveal that the performance of parallel-ism of a single node is much better than that of a master core. Another example is the calculation of solving numeri-cal stress is 80 times faster on the core group than that on the single master core. This experiment could keep a con-stant performance by using 266240 cores, 5.5GFlpos updates of field.

甘霖、钱雪忠、陈宏博

10.12074/202009.00082V1

地球物理学计算技术、计算机技术声学工程

并行计算地震声波正演交错网格高性能计算

甘霖,钱雪忠,陈宏博.基于国产众核处理器三维地震声波正演模拟[EB/OL].(2020-09-28)[2025-08-24].https://chinaxiv.org/abs/202009.00082.点此复制

评论