OpenFAM: Programming disaggregated memory-Reference-Cited by-同舟云学术

OpenFAM: Programming disaggregated memory

Published:2023-09-13 Issue:28 Volume:35 Page:
ISSN:1532-0626
Container-title:Concurrency and Computation: Practice and Experience
language:en
Short-container-title:Concurrency and Computation

Author:

Singhal Sharad¹^ORCID,Crasta Clarete Riana²,Abdulla K Mashood³,Barmawer Faizan³,Bhat Gautham³,Rao Ramya Ahobala³,P N Soumya³,Rajak Rishi Kesh K³

Affiliation:

1. Hewlett Packard Labs Hewlett Packard Enterprise Milpitas California USA

2. HPC Strategy and Operations Hewlett Packard Enterprise New York New York USA

3. HPC Strategy and Operations Hewlett Packard Enterprise Bangalore India

Abstract

AbstractHigh performance computing (HPC) clusters are increasingly handling workloads where working data sets cannot be easily partitioned or are too large to fit into local node memory. In order to enable HPC workloads to access memory external to the node, HPE has defined a programming API (OpenFAM) for developing applications that use large‐scale disaggregated memory. In this paper we describe an open‐source reference implementation of OpenFAM that can be used on scale‐up machines, traditional HPC clusters, as well as emerging disaggregated memory architectures. We demonstrate the efficiency of the implementation using micro‐benchmarks on InfiniBand and Slingshot‐based clusters.

Funder

U.S. Department of Defense

Publisher

Wiley

Subject

Computational Theory and Mathematics,Computer Networks and Communications,Computer Science Applications,Theoretical Computer Science,Software

Link

https://onlinelibrary.wiley.com/doi/am-pdf/10.1002/cpe.7910

Reference38 articles.

1. Ramirez‐GargalloG Garcia‐GasullaM MantovaniF.TensorFlow on state‐of‐the‐art HPC clusters: a machine learning use case. Paper presented at: 2019 19th IEEE/ACM International Symposium on Cluster Cloud and Grid Computing (CCGRID).2019:526‐533.

2. McAlpinJD.Memory Bandwidth and System Balance in HPC Systems Archives;2016.http://sc16.supercomputing.org/tag/memory‐bandwidth‐and‐system‐balance‐in‐hpc‐systems/

3. WeilandM BrunstH QuintinoT et al.An early evaluation of Intel's optane DC persistent memory module and its impact on high‐performance scientific applications. Proceedings of the International Conference for High Performance Computing Networking Storage and Analysis Association for Computing Machinery; New York NY USA.2019:1‐19.

4. DAOS and Intel®.Optane™ Technology for High‐Performance Storage.https://www.intel.com/content/www/us/en/high‐performance‐computing/daos‐high‐performance‐storage‐brief.html

5. SharmaDD.Compute Express Link 2.0 White Paper. tech. rep. Compute Express Link.2020.https://www.computeexpresslink.org/_files/ugd/0c1418_14c5283e7f3e40f9b2955c7d0f60bebe.pdf