Argonne National Laboratory

Exploiting Performance Portability in Search Algorithms for Autotuning

TitleExploiting Performance Portability in Search Algorithms for Autotuning
Publication TypeReport
Year of Publication2015
AuthorsRoy, A, Balaprakash, P, Hovland, PD, Wild, SM
Other NumbersANL/MCS-P5397-0915
Abstract

Autotuning seeks the best configuration of an application by orchestrating hardware and software knobs that affect performance on a given machine. Autotuners adopt various search techniques to efficiently find the best configuration, but they often ignore lessons learned on one machine when tuning for another machine. We demonstrate that a surrogate model built from performance results on one machine can speedup the autotuning search by 1.6X to 130X on a variety of modern architectures.
 

PDFhttp://www.mcs.anl.gov/papers/P5397-0915.pdf