Skip to main content

Exploration of GPGPU register file architecture using domain-wall-shift- write based racetrack memory

Publication ,  Conference
Mao, M; Wen, W; Zhang, Y; Chen, Y; Li, H
Published in: Proceedings - Design Automation Conference
January 1, 2014

SRAM based register le (RF) is one of the major factors lim-iting the scaling of GPGPU. In this work, we propose to use the emerging nonvolatile domain-wall-shift-write based race-track memory (DWSW-RM) to implement a power-effcient GPGPU RF, of which the power consumption is substantially reduced. A holistic technology set is developed to minimize the high access cost of DWSW-RWcaused by the sequential access mechanism. Experiment results show that our proposed tech-niques can improve the GPGPU performance by 4.6% com-pared to the baseline with SRAM based RF. The RF energy effciency is also signcantly improved by 2.45×. Copyright 2014 ACM.

Duke Scholars

Published In

Proceedings - Design Automation Conference

DOI

ISSN

0738-100X

Publication Date

January 1, 2014
 

Citation

APA
Chicago
ICMJE
MLA
NLM
Mao, M., Wen, W., Zhang, Y., Chen, Y., & Li, H. (2014). Exploration of GPGPU register file architecture using domain-wall-shift- write based racetrack memory. In Proceedings - Design Automation Conference. https://doi.org/10.1145/2593069.2593137
Mao, M., W. Wen, Y. Zhang, Y. Chen, and H. Li. “Exploration of GPGPU register file architecture using domain-wall-shift- write based racetrack memory.” In Proceedings - Design Automation Conference, 2014. https://doi.org/10.1145/2593069.2593137.
Mao M, Wen W, Zhang Y, Chen Y, Li H. Exploration of GPGPU register file architecture using domain-wall-shift- write based racetrack memory. In: Proceedings - Design Automation Conference. 2014.
Mao, M., et al. “Exploration of GPGPU register file architecture using domain-wall-shift- write based racetrack memory.” Proceedings - Design Automation Conference, 2014. Scopus, doi:10.1145/2593069.2593137.
Mao M, Wen W, Zhang Y, Chen Y, Li H. Exploration of GPGPU register file architecture using domain-wall-shift- write based racetrack memory. Proceedings - Design Automation Conference. 2014.

Published In

Proceedings - Design Automation Conference

DOI

ISSN

0738-100X

Publication Date

January 1, 2014