Scalable and Fast Lazy Persistency on GPUs

Ardhi Wiratama Baskara Yudha, Keiji Kimura, Huiyang Zhou, Yan Solihin

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

GPUs applications, including many scientific and machine learning applications, increasingly demand larger memory capacity. NVM is promising higher density compared to DRAM and better future scaling potentials. Long running GPU applications can benefit from NVM by exploiting its persistency, allowing crash recovery of data in memory. In this paper, we propose mapping Lazy Persistency (LP) to GPUs and identify the design space of such mapping. We then characterize LP performance on GPUs, varying the checksum type, reduction method, use of locking, and hash table designs. Armed with insights into the performance bottlenecks, we propose a hash table-less method that performs well on hundreds and thousands of threads, achieving persistency with nearly negligible (2.1%) slowdown for a variety of representative benchmarks. We also propose a directive-based programming language support to simplify programming effort for adding LP to GPU applications.

Original languageEnglish
Title of host publicationProceedings - 2020 IEEE International Symposium on Workload Characterization, IISWC 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages252-263
Number of pages12
ISBN (Electronic)9781728176451
DOIs
Publication statusPublished - 2020 Oct
Event16th IEEE International Symposium on Workload Characterization, IISWC 2020 - Virtual, Beijing, China
Duration: 2020 Oct 272020 Oct 29

Publication series

NameProceedings - 2020 IEEE International Symposium on Workload Characterization, IISWC 2020

Conference

Conference16th IEEE International Symposium on Workload Characterization, IISWC 2020
Country/TerritoryChina
CityVirtual, Beijing
Period20/10/2720/10/29

ASJC Scopus subject areas

  • Hardware and Architecture
  • Information Systems and Management

Fingerprint

Dive into the research topics of 'Scalable and Fast Lazy Persistency on GPUs'. Together they form a unique fingerprint.

Cite this