Advanced search
1 file | 877.28 KB

LA-LLC : inter-core locality-aware last-level cache to exploit many-to-many traffic in GPGPUs

Xia Zhao (UGent) , Yuxi Liu (UGent) , Almutaz Adileh (UGent) and Lieven Eeckhout (UGent)
Author
Organization
Abstract
The reply network is a severe performance bottleneck in General Purpose Graphic Processing Units (GPGPUs), as the communication path from memory controllers (MC) to cores is often congested. In this paper, we find that instead of relying on the congested communication path between MCs and cores, the unused core-to-core communication path can be leveraged to transfer data blocks between cores. We propose the inter-core Locality-Aware Last-Level Cache (LA-LLC), which requires only few bits per cache block and enables a core to fetch shared data from another core's private cache instead of the LLC. Leveraging inter-core communication, LA-LLC transforms few-to-many traffic to many-to-many traffic, thereby mitigating the reply network bottleneck. For a set of applications exhibiting varying degrees of inter-core locality, LA-LLC reduces memory access latency and increases performance by 21.1 percent on average and up to 68 percent, with negligible hardware cost.
Keywords
GPGPU, NoC, inter-core locality, LLC

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 877.28 KB

Citation

Please use this url to cite or link to this publication:

Chicago
Zhao, Xia, Yuxi Liu, Almutaz Adileh, and Lieven Eeckhout. 2017. “LA-LLC : Inter-core Locality-aware Last-level Cache to Exploit Many-to-many Traffic in GPGPUs.” Ieee Computer Architecture Letters 16 (1): 42–45.
APA
Zhao, Xia, Liu, Y., Adileh, A., & Eeckhout, L. (2017). LA-LLC : inter-core locality-aware last-level cache to exploit many-to-many traffic in GPGPUs. IEEE COMPUTER ARCHITECTURE LETTERS, 16(1), 42–45.
Vancouver
1.
Zhao X, Liu Y, Adileh A, Eeckhout L. LA-LLC : inter-core locality-aware last-level cache to exploit many-to-many traffic in GPGPUs. IEEE COMPUTER ARCHITECTURE LETTERS. Los alamitos: Ieee Computer Soc; 2017;16(1):42–5.
MLA
Zhao, Xia et al. “LA-LLC : Inter-core Locality-aware Last-level Cache to Exploit Many-to-many Traffic in GPGPUs.” IEEE COMPUTER ARCHITECTURE LETTERS 16.1 (2017): 42–45. Print.
@article{8542501,
  abstract     = {The reply network is a severe performance bottleneck in General Purpose Graphic Processing Units (GPGPUs), as the communication path from memory controllers (MC) to cores is often congested. In this paper, we find that instead of relying on the congested communication path between MCs and cores, the unused core-to-core communication path can be leveraged to transfer data blocks between cores. We propose the inter-core Locality-Aware Last-Level Cache (LA-LLC), which requires only few bits per cache block and enables a core to fetch shared data from another core's private cache instead of the LLC. Leveraging inter-core communication, LA-LLC transforms few-to-many traffic to many-to-many traffic, thereby mitigating the reply network bottleneck. For a set of applications exhibiting varying degrees of inter-core locality, LA-LLC reduces memory access latency and increases performance by 21.1 percent on average and up to 68 percent, with negligible hardware cost.},
  author       = {Zhao, Xia and Liu, Yuxi and Adileh, Almutaz and Eeckhout, Lieven},
  issn         = {1556-6056},
  journal      = {IEEE COMPUTER ARCHITECTURE LETTERS},
  keywords     = {GPGPU,NoC,inter-core locality,LLC},
  language     = {eng},
  number       = {1},
  pages        = {42--45},
  publisher    = {Ieee Computer Soc},
  title        = {LA-LLC : inter-core locality-aware last-level cache to exploit many-to-many traffic in GPGPUs},
  url          = {http://dx.doi.org/10.1109/LCA.2016.2611663},
  volume       = {16},
  year         = {2017},
}

Altmetric
View in Altmetric
Web of Science
Times cited: