Nov 20, 2008

Yi and Whaley receive $462k NSF grant

Qing Yi and Clint Whaley, both assistant professors in the Department of Computer Science, have received a three-year $462,000 grant from the National Science Foundation (NSF) to support their research on "Programmable Code Optimization and Empirical Tuning for High-End Computing." Collaborating with Daniel Quinlan (from Lawrence Livermore National Laboratory), they aim at addressing the complexity facing scientists when developing and tuning high-performance applications.

This project, led by Dr. Yi, will develop a unified framework that allows scientific applications to be tuned for better performance at different stages of the transformation of high-level source code into machine code running on a specific computer system. At the highest level, the applications developer can interactively modify high-level optimization strategies.Then, at an intermediate level, a high-performance computing specialist can specify new program optimizations using a special purpose program transformation language. Finally, at the lowest-level, architecture-specific optimizations can be automatically tuned using empirical performance measurements from the computing system on which the application will run. According to Dr. Yi this separation of concerns in an integrated framework will provide better efficiency, flexibility, and portability than conventional approaches to high-performance computing.The tools implementing this framework will be released as open-source software.

Interested parties should contact Dr. Yi [] for more information.