The Tuning and Analysis Utilities (TAU) project from the University of Oregon provides a performance system framework for scalable parallel and distributed high-performance computing. TAU is capable of various performance instrumentation, measurement and analysis tasks, many with visualization components. It integrates with other, lower-level tools (such as PAPI) where appropriate. TAU is a suite of tools that were built to help developers identify the hotspots in their parallel code so that they can focus optimization efforts on the rate-limiting fraction of their code.

 
©   Cornell University  |  Center for Advanced Computing  |  Copyright Statement  |  Inclusivity Statement