If you were hoping for a glimpse into the National Strategic Computing Initiative implementation (NSCI) plan last week while at SC15, you were out of luck. A draft, said to be quite long, has been completed but is being closely guarded at the moment. Perhaps in February, suggested William Polk of the Office of Science and Technology, closer to the final phases of federal budget preparation we’ll finally see the plan.
Currently NSCI hopes are high but questions persist. Until the implementation plan is made public, much remains a mystery. The outreach to the HPC community, of course, has been ongoing. Of note, the RFI promulgated by NSF produced roughly 200 responses, but only eight from industry, Polk told the audience during an NSCI update he delivered at SC15 with colleague Randy Bryant also of OTSP.
While at the conference William Gropp and Thomas Sterling, distinguished HPC leaders (affiliation below) and co-editors of HPCwire’s Exascale Edition, generously took time to discuss NSCI with HPCwire managing editor, John Russell. The conversation ranged over many topics such as: what Gropp and Sterling hope to see in the implementation plans; which technologies are the most pressing; how best to mount a broad computing community effort; and the need, perhaps, to fundamentally change the computing paradigm.
Dr. Gropp is a professor of computer science at the University of Illinois Urbana-Champaign, where, since 2013, he has held the Thomas M. Siebel Chair. Dr. Sterling is professor of intelligent systems engineering at the Indiana University (IU) School of Informatics and Computing as well as chief scientist and associate director of the Center for Research in Extreme Scale Technologies (CREST).