Researchers tackle resiliency issues in exascale computing

11/26/2012 |

The built-in complexity that makes supercomputers so powerful also makes them more prone to complications, according to researchers, who say the hundreds of thousands of components in high-performance computing systems make them vulnerable to resiliency issues. A group led by David Fiala, a doctoral student at North Carolina State University, is pioneering advanced redundancy technology for exascale computing systems known as RedMPI, designed to address silent data corruption by simultaneously running multiple program "clones."

View Full Article in:

Published in Brief: