Computational Sprinting: Exceeding Sustainable Power in Thermally Constrained Systems.Presented at the 18th Symposium on High Performance Computer Architecture (HPCA), February 2012 MartinДў012 Dark Silicon Workshop (in conjunction with ISCA 2012) Computational Sprinting on a Real System: Preliminary ResultsДЄrun Raghavan, Marios Papaefthymiou, Kevin P.Computational Sprinting Overview Talk - June 2012.Presented at the 18th Eighteenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), March 2013. Computational Sprinting on a Hardware/Software Testbed.Presented at the Workshop on Duplicating, Deconstructing, and Debunking (WDDD), June 2014. In the Proceedings of the 18th Symposium on High Performance Computer Architecture (HPCA), February 2012 Computational Sprinting (HPCA best paper award).In the Proceedings of the 18th Eighteenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), March 2013. Computational Sprinting on a Hardware/Software Testbed (ASPLOS best paper award).Designing for Responsiveness with Computational Sprinting (IEEE Micro 'Top Picks')ДЄrun Raghavan, Yixin Luo, Anuj Chandawalla, Marios Papaefthymiou, Kevin P.IEEE Micro Special Issue on Dark Silicon, IEEE Micro, Volume 33, Number 5, Sep-Oct 2013 Utilizing Dark Silicon to Save Energy with Computational SprintingДЄrun Raghavan, Laurel Emurian, Lei Shao, Marios Papaefthymiou, Kevin P.Proceedings of the 30th Annual Thermal Measurement, Modeling, and Management Symposium (SemiTherm), Mar 2014 Lei Shao, Arun Raghavan, Laurel Emurian, Marios Papaefthymiou, Thomas F. On-chip Phase Change Heat Sinks Designed for Computational Sprinting.Workshop on Duplicating, Deconstructing, and Debunking (WDDD), June 2014 Laurel Emurian, Arun Raghavan, Lei Shao, Jeffrey M. Pitfalls of Accurately Benchmarking Thermally Adaptive Chips.To effectively facilitate sprinting for short time durations overcoming the physicalД¬hallenges inherent in our target environments. This project therfore explores various thermal, electrical, architectural and software/runtime aspects (power delivery and stability) capacities of a typical smart-phone like device. Punctuated by long idle periods waiting for user input.Д¬omputational sprinting activates otherwise powered-down cores for sub-secondД«ursts of intense parallel computation in response to such sporadic user activity.Дuring sprints, the processor generates heat at a rate that far exceeds the thermal (cooling) and electrical Where many interactive applications are characterized by short bursts of computational demand Our approach, called computational sprinting is aimed at mobile environments like smart-phones, We pose the question: 'What would a system look like if designed to provide responsivenessДuring bursts rather than with a singular focus on sustained performance?' Heat sinks) are designed primarly for sustained performance, While conventional processor designs (including their energy delivery systems and

The long dark sprinting