Computing power to answer NASA's complex science and engineering questions
HECC offers a range of consulting services for optimizing your application performance and productivity, in order to facilitate advances in science and engineering for NASA missions.
Our application performance and productivity (APP) experts work closely with application owners and domain specialists to make the most effective use of the high-performance computing (HPC) resources by enhancing code performance—often by an order of magnitude.
To request any of the following optimization services at no charge, please contact the NAS Control Room: (800) 331-8737, (650) 604-4444, support@nas.nasa.gov
Our optimization experts are also available to work closely with the code owners and domain experts to conduct in-depth application analysis, and potentially undertake major code re-engineering to enhance and optimize code performance. Such services require external funding based on the required level of effort.
For further information on these in-depth services, please contact:
Piyush Mehrotra
Application Performance & Productivity Group Lead
piyush.mehrotra@nasa.gov
(650) 604-5126
It's surprising how small code adjustments can make a large impact on code performance. Whether making simple or comprehensive code changes, the APP group's work has greatly benefited research applications for dozens of projects in all NASA mission areas. For example, the team reduced the USM3D code's runtime and memory requirements by factor of 2 to support very large computational grids. USM3D is an important computational fluid dynamics (CFD) code used for intensive aerodynamic analysis.
Can't find what you're looking for? NAS Control Room staff are available 24x7x365: (800) 331-8737, (650) 604-4444, support@nas.nasa.gov
Optimization Yields 2x Speedup for US3D Hypersonics Modeling Code
Optimization of the US3D CFD code resulted in a 2x speedup for the code overall when executing a standard real-world test case of over 30 million grid elements on the Pleiades supercomputer.
Read More