NUMA-Aware Data-Transfer Measurements for Power/NVLink Multi-GPU Systems

Carl Pearson, I. Hsin Chung, Zehra Sura, Wen Mei Hwu, Jinjun Xiong

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


High-performance computing increasingly relies on heterogeneous systems with specialized hardware accelerators to improve application performance. For example, NVIDIA’s CUDA programming system and general-purpose GPUs have emerged as a widespread accelerator in HPC systems. This trend has exacerbated challenges of data placement as accelerators often have fast local memories to fuel their computational demands, but slower interconnects to feed those memories. Crucially, real-world data-transfer performance is strongly influenced not just by the underlying hardware, but by the capabilities of the programming systems. Understanding how application performance is affected by the logical communication exposed through abstractions, as well as the underlying system topology, is crucial for developing high-performance applications and architectures. This report presents initial data-transfer microbenchmark results from two POWER-based systems obtained during work towards developing an automated system performance characterization tool.

Original languageEnglish (US)
Title of host publicationHigh Performance Computing - ISC High Performance 2018 International Workshops, Revised Selected Papers
EditorsMichèle Weiland, Rio Yokota, Sadaf Alam, John Shalf
PublisherSpringer-Verlag Berlin Heidelberg
Number of pages7
ISBN (Print)9783030024642
StatePublished - 2018
EventInternational Conference on High Performance Computing, ISC High Performance 2018 - Frankfurt, Germany
Duration: Jun 28 2018Jun 28 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11203 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


ConferenceInternational Conference on High Performance Computing, ISC High Performance 2018


  • Benchmark
  • CUDA
  • NVLink
  • Unified Memory

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'NUMA-Aware Data-Transfer Measurements for Power/NVLink Multi-GPU Systems'. Together they form a unique fingerprint.

Cite this