Solving Poisson’s equation using FFT in a GPU cluster

详细信息    查看全文

• 作者：Jose L. Jodra^a ; ^{joseluis.jodra@ehu.eus}Author Vitae ; Ibai Gurrutxaga^b ; ^{i.gurrutxaga@ehu.eus}Author Vitae ; Javier Muguerza^b ; ^{j.muguerza@ehu.eus}Author Vitae ; Ainhoa Yera^b ; ^{ainhoa.yera@ehu.eus}Author Vitae
• 关键词：GPU ; Poisson&rsquo ; s equation ; MPI ; FFT ; CUDA
• 刊名：Journal of Parallel and Distributed Computing
• 出版年：2017
• 出版时间：April 2017
• 年：2017
• 卷：102
• 期：Complete
• 页码：28-36
• 全文大小：839 K
• 卷排序：102

文摘

We implement the solution of Poisson’s equation using FFT in a cluster of GPUs. We overlap computation, inter-node communication and CPU/GPU communication. The results show good scalability up to 16 GPUs. Our implementation is about 2.5 times faster than an optimized CPU based solver.