By Todd Munson
A vibrant, high-quality software ecosystem is essential to realizing the full potential of exascale computing resources in helping to solve some of the world’s greatest scientific challenges. The Exascale Computing Project (ECP) has brought the scientific software community together to address various types of software-related issues, such as quality and sustainability.
As we continue to exploit the power of exascale computing and advance our understanding, we have the opportunity to reimagine how our software technologies should be developed and deployed, as we re-architect them to better support the unique features of these machines.
A greater emphasis on software architecture and sustainability will put us on a trajectory to better meet the needs of many future generations of computing resources. Moreover, packaging software technologies into software development kits will allow application developers to easily adopt our scientific libraries and have confidence in both their quality and their ability to run at scale on the exascale computing platforms when they are delivered.
Future investments in research, development, delivery, and deployment activities for our software technologies will produce broad benefits that we can only imagine today as we move into the exascale era and beyond.