I love it that we have this great software pathfinding team looking for new ways to make it easier to tune your Parallel code for multi-core systems. The team has recently posted a new performance tuning tool, the Intel Software Autotuning Tool (ISAT) that supports automatic searching for the near-optimal values of program parameters that have significant performance impact. Think of such examples as the cache-blocking factors in matrix computations, the task granularity in Threading Building Blocks, and the scheduling policy of OpenMP parallel constructs. The output is a tuned version of the program based on the searching results. This tool is easy to use with no compiler dependency. You can learn more and download the Intel Software Autotuning Tool on Intel’s Parallel Programming Community. Check it out and let us know how it performs for you.
