SDSC Computing the 21st Century Talk Given to the NSF Sugar Panel May 27, 1998
SDSC Scientific Applications Continue to Require Exponential Growth in Capacity MACHINE REQUIREMENT IN FLOPS NSF Capability NSF Leading Edge Molecular Dynamics for Biological Molecules Computational Cosmology Turbulent Convection in Stars Atomic/Diatomic Interaction QCD MEMORYMEMORY BYTES BYTES = Long Range Projections from Recent Applications Workshop = Next Step Projections by NSF Grand Challenge Research Teams = Recent Computations by NSF Grand Challenge Research Teams ASCI in year climate model in hours NSF in 2004 (Projected) From Bob Voigt, NSF
SDSC The Problem Major Near Term Shortfall in Computing Capacity –NSF PIs have inadequate access to resources –Result of Failure to Transfer PSC T3E NSF PACI is Underfunded to Achieve Both Goals of: –Providing Computational Production Environment –Prototyping 21st Century Infrastructure Limited NSF PI Access to –Visualization and Analysis Facilities –Key Scientific Data Sets ASCI Program is Capturing Academic Leadership
SDSC Let’s Blow This Up! The Growth Rate of the National Capacity is Slowing Down again Source: Quantum Research; Lex Lane, NCSA
SDSC Major Gap Developing in National Usage at NSF Supercomputer Centers 70% Annual Growth Rate is the Historical Rate of National Usage Growth. It is Also Slightly Greater Than the Rate of Moore’s Law, So Lesser Growth Means Desktops Gain on Supers Projection
SDSC Monthly National Usage at NSF Supercomputer Centers FY96FY97FY98FY99 Projection Capacity Level NSF Proposed at 3/97 NSB Meeting Transition Period
SDSC Allocated Capacity for NSF Large National Projects Doubled Data from NSF Metacenter and NRAC Reviews
SDSC Accelerated Strategic Computing Initiative is Coupling DOE DP Labs to Universities Access to ASCI Leading Edge Supercomputers Academic Strategic Alliances Program Data and Visualization Corridors
SDSC Comparison of the DoE ASCI and the NSF PACI Origin Array Scale Through FY99 /Hardware/schedule.html Los Alamos Origin System FY processors NCSA Proposed System FY99 6x128 and 4x64=1024 processors
SDSC DOE Strategic Simulation Initiative Major New FY00 Computational Initiative for DOE ER Large-scale Problems in Energy and the Environment Initial Targets Are Climate Simulation and Combustion Major Investment in Applications, Computer Science & Applied Mathematics, and Computational Facilities Pursued Jointly With Universities, Labs and Industry Will Leverage ASCI Investments Will Field Machines in the 40-50TF Range in 2003 Opportunity for Synergy with NSF PACI Courtesy of Rick Stevens, Argonne
SDSC The Emerging Multi-Agency Effort to Create a Persistent National Technology Grid NSF Working Closely with NPACI & Alliance KDI Proposals Will Require PACI Infrastructure DoE Discussing with NSF –How to Get Highest-End Capacity –PACI ET and AT Teams Linking to ASCI, SSI NASA MOU with NSF on Computational Grids DoD Mod. PET Innovating Infrastructure NIH Potentially Critical Partner
SDSC Possible NSF Initiative Enhance PACI Base Funding –To Link Computer Scientists to Application Teams –Prototype 21st Century Information Infrastructure Augment PACI Base Funding by Special Initiative –Fund a Backbone Production “Digital Fabric” –Augment Regional and Local Facilities Additional FY00 Funds –From Foundation-Wide Approach –Possible Funding from Other Agencies Create Foundation-Wide Advisory Board –Computational Scientists and Engineers –PIs of NSF ERC, STC, KDI, National Observatories, DLI-II
SDSC Proposed National Computational Science, Engineering, and Knowledge Management Facility Advanced Visualization and Analysis Facilities Distributed Datasets and Digital Libraries Leading Edge Scalable Computation Facilities Academic User Community Networking