License: Creative Commons Attribution 3.0 Unported license (CC BY 3.0)
When quoting this document, please refer to the following
DOI: 10.4230/LIPIcs.ECRTS.2018.19
URN: urn:nbn:de:0030-drops-89833
URL: http://dagstuhl.sunsite.rwth-aachen.de/volltexte/2018/8983/
Go to the corresponding LIPIcs Volume Portal

Ali, Waqar ; Yun, Heechul

Protecting Real-Time GPU Kernels on Integrated CPU-GPU SoC Platforms

pdf-format:
LIPIcs-ECRTS-2018-19.pdf (0.9 MB)

Abstract

Integrated CPU-GPU architecture provides excellent acceleration capabilities for data parallel applications on embedded platforms while meeting the size, weight and power (SWaP) requirements. However, sharing of main memory between CPU applications and GPU kernels can severely affect the execution of GPU kernels and diminish the performance gain provided by GPU. For example, in the NVIDIA Jetson TX2 platform, an integrated CPU-GPU architecture, we observed that, in the worst case, the GPU kernels can suffer as much as 3X slowdown in the presence of co-running memory intensive CPU applications. In this paper, we propose a software mechanism, which we call BWLOCK++, to protect the performance of GPU kernels from co-scheduled memory intensive CPU applications.

BibTeX - Entry

@InProceedings{ali_et_al:LIPIcs:2018:8983,
  author =	{Waqar Ali and Heechul Yun},
  title =	{{Protecting Real-Time GPU Kernels on Integrated CPU-GPU SoC Platforms}},
  booktitle =	{30th Euromicro Conference on Real-Time Systems (ECRTS 2018)},
  pages =	{19:1--19:22},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-075-0},
  ISSN =	{1868-8969},
  year =	{2018},
  volume =	{106},
  editor =	{Sebastian Altmeyer},
  publisher =	{Schloss Dagstuhl--Leibniz-Zentrum fuer Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{http://drops.dagstuhl.de/opus/volltexte/2018/8983},
  URN =		{urn:nbn:de:0030-drops-89833},
  doi =		{10.4230/LIPIcs.ECRTS.2018.19},
  annote =	{Keywords: GPU, memory bandwidth, resource contention, CPU throttling, fair scheduler}
}

Keywords: GPU, memory bandwidth, resource contention, CPU throttling, fair scheduler
Collection: 30th Euromicro Conference on Real-Time Systems (ECRTS 2018)
Issue Date: 2018
Date of publication: 22.06.2018
Supplementary Material: ECRTS Artifact Evaluation approved artifact available at https://dx.doi.org/10.4230/DARTS.4.2.3

DROPS-Home | Fulltext Search | Imprint | Privacy Published by LZI