The READEX tool suite supports users to improve the energy-efficiency of their HPC applications. It enables them to exploit the dynamic behaviour of their applications by adjusting the system to the actual resource requirements. More detailed information are available in the Project Overview. If you want to have a look on our sources, check our Github account. You can also download READEX at our download page.

Results

We tested the READEX tool suite on different applications and benchmarks, parallelized with OpenMP and/or MPI. The following picture shows the percentage and absolute energy savings for applying hardware and runtime tuning parameters.

Overview

How to use

 News

• Release Candidate Published! - Here it is! The release candidate can now be downloaded at our download page. It contains some final fixes that increase user experience. Please, also checkout the new metric plugins, […]
• Pre-release candidate - For those of you, who missed it: We updated our software tar balls to the pre-release candidate on August, 1st. Edit: To avoid downloading the wrong READEX version, we removed […]
• READEX Beta 1806 - Within the last month we released beta versions at regular intervals. Yesterday, we uploaded the last beta version of the READEX tool suite, which will be followed by the pre-release […]
• EETHPC at ISC 2018 - And another workshop. During the ISC High Performance 2018, we will co-host the workshop Energy Efficiency Tools for High Performance Computing (EETHPC). Check out the program at the workshop website. […]
• HPCAFE at European HPC Summit Week 2018 - Like last week, the READEX project together with the AllScale, the ANTAREX, and the EXCAPE project is organising the High-Performance Computing Approaches for Monitoring, Exploring, Optimizing and Autotuning (HPCAFE-2018) as […]