Prospector: Synthesizing Efficient Accelerators via Statistical Learning
Publication
, Conference
Mehrabi, A; Manocha, A; Lee, BC; Sorin, DJ
Published in: Proceedings of the 2020 Design, Automation and Test in Europe Conference and Exhibition, DATE 2020
March 1, 2020
Accelerator design is expensive due to the effort required to understand an algorithm and optimize the design. Architects have embraced two technologies to reduce costs. High-level synthesis automatically generates hardware from code. Reconfigurable fabrics instantiate accelerators while avoiding fabrication costs for custom circuits. We further reduce design effort with statistical learning. We build an automated framework, called Prospector, that uses Bayesian techniques to optimize synthesis directives, reducing execution latency and resource usage in field-programmable gate arrays. We show in a certain amount of time designs discovered by Prospector are closer to Pareto-efficient designs compared to prior approaches.
Duke Scholars
Published In
Proceedings of the 2020 Design, Automation and Test in Europe Conference and Exhibition, DATE 2020
DOI
ISBN
9783981926347
Publication Date
March 1, 2020
Start / End Page
151 / 156
Citation
APA
Chicago
ICMJE
MLA
NLM
Mehrabi, A., Manocha, A., Lee, B. C., & Sorin, D. J. (2020). Prospector: Synthesizing Efficient Accelerators via Statistical Learning. In Proceedings of the 2020 Design, Automation and Test in Europe Conference and Exhibition, DATE 2020 (pp. 151–156). https://doi.org/10.23919/DATE48585.2020.9116473
Mehrabi, A., A. Manocha, B. C. Lee, and D. J. Sorin. “Prospector: Synthesizing Efficient Accelerators via Statistical Learning.” In Proceedings of the 2020 Design, Automation and Test in Europe Conference and Exhibition, DATE 2020, 151–56, 2020. https://doi.org/10.23919/DATE48585.2020.9116473.
Mehrabi A, Manocha A, Lee BC, Sorin DJ. Prospector: Synthesizing Efficient Accelerators via Statistical Learning. In: Proceedings of the 2020 Design, Automation and Test in Europe Conference and Exhibition, DATE 2020. 2020. p. 151–6.
Mehrabi, A., et al. “Prospector: Synthesizing Efficient Accelerators via Statistical Learning.” Proceedings of the 2020 Design, Automation and Test in Europe Conference and Exhibition, DATE 2020, 2020, pp. 151–56. Scopus, doi:10.23919/DATE48585.2020.9116473.
Mehrabi A, Manocha A, Lee BC, Sorin DJ. Prospector: Synthesizing Efficient Accelerators via Statistical Learning. Proceedings of the 2020 Design, Automation and Test in Europe Conference and Exhibition, DATE 2020. 2020. p. 151–156.
Published In
Proceedings of the 2020 Design, Automation and Test in Europe Conference and Exhibition, DATE 2020
DOI
ISBN
9783981926347
Publication Date
March 1, 2020
Start / End Page
151 / 156