GPU acceleration of the Locally Selfconsistent Multiple Scattering code for first principles calculation of the ground state and statistical physics of materials

详细信息    查看全文

• 作者：Markus Eisenbach^a ; ^{eisenbachm@ornl.gov} ; Jeff Larkin^b ; Justin Lutjens^b ; Steven Rennich^b ; James H. Rogers^a
• 关键词：First-principles ; Monte-Carlo ; Phase transitions
• 刊名：Computer Physics Communications
• 出版年：2017
• 出版时间：February 2017
• 年：2017
• 卷：211
• 期：Complete
• 页码：2-7
• 全文大小：748 K
• 卷排序：211

文摘

The Locally Self-consistent Multiple Scattering (LSMS) code solves the first principles Density Functional theory Kohn–Sham equation for a wide range of materials with a special focus on metals, alloys and metallic nano-structures. It has traditionally exhibited near perfect scalability on massively parallel high performance computer architectures. We present our efforts to exploit GPUs to accelerate the LSMS code to enable first principles calculations of O(100,000) atoms and statistical physics sampling of finite temperature properties. We reimplement the scattering matrix calculation for GPUs with a block matrix inversion algorithm that only uses accelerator memory. Using the Cray XK7 system Titan at the Oak Ridge Leadership Computing Facility we achieve a sustained performance of 14.5PFlop/s and a speedup of 8.6 compared to the CPU only code.