We present an overview of current research in performance modeling, focusing on efforts underway in the Performance Evaluation Research Center (PERC). Using some new techniques, we are able to construct performance models that can be used to project the sustained performance of large-scale scientific programs on different systems, over a range of job and system sizes. Such models can be used by vendors in system designs, by computing centers in system acquisitions, and by application scientists to improve the performance of their codes.
David H. Bailey, Allan Snavely