BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20111118T013000Z
DTEND:20111118T030000Z
LOCATION:WSCC 3A
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: Until the proposed U.S. Department of Energy Exascale Initiative becomes official, the path and approach of the National Nuclear Security Administration’s Advanced Simulation and Computing program for the next decade will be "evolutionary": how we might bend or adapt existing technologies to serve as a basis and environment for computing at Exascale?  Many believe that such an approach is sub-optimal, that the unique problems or unprecedented scale represent an opportunity to reexamine past choices, or that the "evolutionary" approach will introduce significant software engineering risks, without complementary mitigation. In this interactive panel, composed of DOE/NNSA researchers, we will present some thoughts on alternative out-of-the-box or "revolutionary" options, actively solicit ideas from the audience, and moderate a group discussion between the panel and audience on the topic. The scope is broad, encompassing programming environments, altered application strategies, architecture, I/O. on-system and real-time data triage and analysis, power management, and operations.
SUMMARY:Punctuated Equilibrium at Exascale
PRIORITY:3
END:VEVENT
END:VCALENDAR