Optimizing Amber for Device-to-Device GPU Communication

Optimizing Amber for Device-to-Device GPU Communication
复制标题

优化 Amber 以实现设备间 GPU 通信

DOI:
10.1145/3569951.3597553
复制
发表时间:
2023
期刊:
ACM
影响因子:
--
通讯作者:
Panda, Dhabaleswar K.
Panda, Dhabaleswar K.
中科院分区:
--
文献类型:
--
作者:
Khuvis, Samuel;Tomko, Karen;Brozell, Scott R.;Chen, Chen-Chun;Subramoni, Hari;Panda, Dhabaleswar K.

文献摘要

参考文献

相似文献

尽管GPU到GPU的直接通信在MPI库中已经实现了十多年,但学术HPC中心兼容硬件的有限可用性阻碍了利用此功能的科学应用程序中算法的开发。在本文中,我们采取琥珀,用于模拟蛋白质和核酸的分子动力学代码,作为一个测试案例。我们演示了实现GPU到GPU通信所需的修改。与以前的实现相比,这些修改显示出平均约36%的整体性能提高和84%的重要显式溶剂子集的基准。
Although direct GPU-to-GPU communication has been possible in MPI libraries for over a decade, the limited availability of compatible hardware at academic HPC centers has discouraged the development of algorithms in scientific applications that take advantage of this capability. In this paper, we take Amber, a molecular dynamics code used to simulate proteins and nucleic acids, as a test case. We demonstrate the modifications necessary to implement GPU-to-GPU communication. Compared to the previous implementation, these modifications show an average of approximately 36% improvement in performance overall and 84% for the important explicit solvent subset of the benchmarks.
DOI: 10.1146/annurev-biophys-091720-102019
发表时间: 2021-05-06
影响因子: 12.4
作者:
Schlick T;Portillo-Ledesma S;Myers CG;Beljak L;Chen J;Dakhel S;Darling D;Ghosh S;Hall J;Jan M;Liang E;Saju S;Vohr M;Wu C;Xu Y;Xue E
通讯作者: Xue E
MVAPICH 项目:将研究转化为 HPC 社区的高性能 MPI 库
DOI: 10.1016/j.jocs.2020.101208
发表时间: 2020
影响因子: 3.3
作者:
Panda, Dhabaleswar Kumar;Subramoni, Hari;Chu, Ching-Hsiang;Bayatpour, Mohammadreza
通讯作者: Bayatpour, Mohammadreza
AMBER 免费能源工具:优化炼金转化途径设计的新框架。
DOI: 10.1021/acs.jctc.2c00725
发表时间: 2023
影响因子: 5.5
作者:
Tsai,Hsu-Chun;Lee,Tai-Sung;Ganguly,Abir;Giese,TimothyJ;Ebert,MaximilianCcjc;Labute,Paul;MerzJr,KennethM;York,DarrinM
通讯作者: York,DarrinM