J. Chame, C. Chen, J. Dongarra, M. Hall, J. K. Hollingsworth, P. Hovland, S. Moore, K. Seymour, J. Shin, A. Tiwari, S. Williams, H. You, and D. H. Bailey
Preprint Version: [pdf]
The enormous and growing complexity of today's high-end systems has
increased the already significant challenges of obtaining high performance on today's equally complex scientific applications. Application scientists are faced with a daunting challenge in tuning their codes to exploit performance-enhancing architectural features. The Performance Engineering Research Institute (PERI) is working towards the goal of automating portions of the performance tuning process. This paper describes PERI’s overall strategy for auto-tuning tools, and recent progress in both building auto-tuning tools and demonstrating
their success on kernels, some taken from large-scale applications.