The Community for Technology Leaders
RSS Icon
Issue No.05 - Sept.-Oct. (2013 vol.33)
pp: 20-28
Arun Raghavan , University of Pennsylvania
Laurel Emurian , University of Pennsylvania
Lei Shao , University of Michigan
Marios Papaefthymiou , University of Michigan
Kevin P. Pipe , University of Michigan
Thomas F. Wenisch , University of Michigan
Milo M. K. Martin , University of Pennsylvania
Computational sprinting activates dark silicon to improve responsiveness by briefly but intensely exceeding a system's sustainable power limit. This article focuses on the energy implications of sprinting. The authors observe that sprinting can save energy even while improving responsiveness by enabling execution in chip configurations that, though thermally unsustainable, improve energy efficiency. Surprisingly, this energy savings can translate to throughput improvements even for long-running computations. Repeatedly alternating between sprint and idle modes while maintaining sustainable average power can outperform steady-state computation at the platform's thermal limit.
Silicon, Time-frequency analysis, Semiconductor device manufacture, Temperature measurement, Computational modeling, Analytical models, Semiconductor device manufacture, Energy efficiency,responsiveness, dark silicon, energy-aware architecture, thermal-aware architecture, temperature
Arun Raghavan, Laurel Emurian, Lei Shao, Marios Papaefthymiou, Kevin P. Pipe, Thomas F. Wenisch, Milo M. K. Martin, "Utilizing Dark Silicon to Save Energy with Computational Sprinting", IEEE Micro, vol.33, no. 5, pp. 20-28, Sept.-Oct. 2013, doi:10.1109/MM.2013.76
31 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool