Exploring new features of high-bandwidth memory for GPUs

Bingchao Li, Choungki Song, Jizeng Wei, Jung Ho Ahn, Nam Sung Kim

Research output: Contribution to journalArticle

Abstract

Due to the off-chip I/O pin and power constraints of GDDR5, HBM has been proposed to provide higher bandwidth and lower power consumption for GPUs. In this paper, we first provide detailed comparison between HBM and GDDR5 and expose two unique features of HBM: Dualcommand and pseudo channel mode. Second, we analyze the effectiveness of these two features and show that neither notably contributes to performance. However, by combining pseudo channel mode with cache architecture supporting fine-grained cache-line management such as Amoeba cache, we achieve high effciency for applications with irregular memory requests. Our experiment demonstrates that compared with Amoeba caches with legacy mode, Amoeba cache with pseudo channel mode improves GPU performance by 25% and reduces HBM energy consumption by 15%.

Original languageEnglish (US)
Article number20160527
JournalIEICE Electronics Express
Volume13
Issue number14
DOIs
StatePublished - Jan 1 2016

Keywords

  • Amoeba cache
  • DRAM
  • Energy
  • GPU
  • HBM

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Exploring new features of high-bandwidth memory for GPUs'. Together they form a unique fingerprint.

  • Cite this

    Li, B., Song, C., Wei, J., Ahn, J. H., & Kim, N. S. (2016). Exploring new features of high-bandwidth memory for GPUs. IEICE Electronics Express, 13(14), [20160527]. https://doi.org/10.1587/elex.13.20160527