Performance-Portable Numerical Relativity with AthenaK

arXiv - PHYS - General Relativity and Quantum Cosmology Pub Date : 2024-09-16 DOI:arxiv-2409.10383

Hengrui Zhu, Jacob Fields, Francesco Zappa, David Radice, James Stone, Alireza Rashti, William Cook, Sebastiano Bernuzzi, Boris Daszuta

{"title":"Performance-Portable Numerical Relativity with AthenaK","authors":"Hengrui Zhu, Jacob Fields, Francesco Zappa, David Radice, James Stone, Alireza Rashti, William Cook, Sebastiano Bernuzzi, Boris Daszuta","doi":"arxiv-2409.10383","DOIUrl":null,"url":null,"abstract":"We present the numerical relativity module within AthenaK, an open source\nperformance-portable astrophysics code designed for exascale computing\napplications. This module employs the Z4c formulation to solve the Einstein\nequations. We demonstrate its accuracy through a series of standard numerical\nrelativity tests, including convergence of the gravitational waveform from\nbinary black hole coalescence. Furthermore, we conduct scaling tests on OLCF\nFrontier and NERSC Perlmutter, where AthenaK exhibits excellent weak scaling\nefficiency of 80% on up to 65,536 AMD MI250X GPUs on Frontier (relative to 4\nGPUs) and strong scaling efficiencies of 84% and 77% on AMD MI250X and NVIDIA\nA100 GPUs on Frontier and Perlmutter respectively. Additionally, we observe a\nsignificant performance boost, with two orders of magnitude speedup ($\\gtrsim\n200\\times$) on a GPU compared to a single CPU core, affirming that AthenaK is\nwell-suited for exascale computing, thereby expanding the potential for\nbreakthroughs in numerical relativity research.","PeriodicalId":501041,"journal":{"name":"arXiv - PHYS - General Relativity and Quantum Cosmology","volume":"52 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - General Relativity and Quantum Cosmology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.10383","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

Abstract

We present the numerical relativity module within AthenaK, an open source performance-portable astrophysics code designed for exascale computing applications. This module employs the Z4c formulation to solve the Einstein equations. We demonstrate its accuracy through a series of standard numerical relativity tests, including convergence of the gravitational waveform from binary black hole coalescence. Furthermore, we conduct scaling tests on OLCF Frontier and NERSC Perlmutter, where AthenaK exhibits excellent weak scaling efficiency of 80% on up to 65,536 AMD MI250X GPUs on Frontier (relative to 4 GPUs) and strong scaling efficiencies of 84% and 77% on AMD MI250X and NVIDIA A100 GPUs on Frontier and Perlmutter respectively. Additionally, we observe a significant performance boost, with two orders of magnitude speedup ($\gtrsim 200\times$) on a GPU compared to a single CPU core, affirming that AthenaK is well-suited for exascale computing, thereby expanding the potential for breakthroughs in numerical relativity research.

查看原文本刊更多论文

利用 AthenaK 实现性能便携的数值相对论

我们介绍了 AthenaK 中的数值相对论模块，AthenaK 是为超大规模计算应用而设计的开源高性能可移植天体物理学代码。该模块采用 Z4c 公式求解爱因斯坦方程。我们通过一系列标准数值相对论测试证明了它的准确性，包括双黑洞凝聚产生的引力波形的收敛性。此外，我们还在OLCFFrontier和NERSC Perlmutter上进行了扩展测试，结果表明AthenaK在Frontier上高达65,536个AMD MI250X GPU（相对于4GPU）上表现出80%的出色弱扩展效率，而在Frontier和Perlmutter上的AMD MI250X和NVIDIAA100 GPU上分别表现出84%和77%的强扩展效率。此外，我们还观察到显著的性能提升，与单个CPU内核相比，GPU上的速度提升了两个数量级（$\gtrsim200\times$），这肯定了AthenaK非常适合超大规模计算，从而扩大了数值相对论研究取得突破的潜力。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

arXiv - PHYS - General Relativity and Quantum Cosmology

自引率

0.00%

发文量