NSF Vision and Strategy for Advanced Computational Infrastructure Vision: NSF Leadership in creating and deploying a comprehensive portfolio…to facilitate cutting – edge foundational research in computational and data-enabled science and engineering…. Five Strategies: Foundational research to fully exploit parallelism and concurrency…. Applications research and development…. Building, testing, and deploying both sustainable and innovative resources into a collaboration ecosystem…. Comprehensive education and workforce programs… Development and evaluation of transformational and grand challenge community programs… Image Credit: NCSA/University of Illinois Blue Waters Image Credit: University of Tennessee, Knoxville Image Credit: TACC Stampede Image Credit: UC San Diego Publications/Erik Jepsen Gordon

Major NSF Advanced Computing Infrastructure deployed in FY 2013 NCAR/ Wyoming Supercomputing Center Image Credit: TACC Stampede, UT Austin Image Credits: NCSA/University of Illinois Blue Waters, UIUC Image Credit: NCAR - Wyoming Supercomputing Center

Allocated thru XSEDE Forthcoming Resources * To be retired in 2014 Resources are increasingly diverse and collaborative Blacklight* Shared Memory 4k Xeon Cores Blacklight* Shared Memory 4k Xeon Cores Kraken* 110K Cores Keeneland* CPU/GPGPU Kraken* 110K Cores Keeneland* CPU/GPGPU Stampede 460K Cores w. Xeon Phi >1000 Users Lonestar* Large Memory Stampede 460K Cores w. Xeon Phi >1000 Users Lonestar* Large Memory Maverick Visualization Data Analytics Wrangler Date Intensive Trestles IO-Intensive 10k Cores 160 GB SSD/Flash Gordon Data Intensive 64 TB Memory 300 TB Flash Memory Trestles IO-Intensive 10k Cores 160 GB SSD/Flash Gordon Data Intensive 64 TB Memory 300 TB Flash Memory FutureGrid* CS testbed Open Science Grid High Throughput Blue Waters Leadership Class Comet “Long Tail Science” 47k Cores/2 PF SuperMIC 380 nodes – 1PF (Ivy bridge, Xeon Phi, GPU) Yellowstone Geosciences Image credit: Pittsburgh Supercomputing Center Image Credit: Alan Decker, SDSC

Trends: Within TeraGrid and successor XSEDE, the Number of Research Projects, Institutions and Users have increased

While the number of resources has decreased, actual core hours increased but requests remain 2-5x available

All Scientific Disciplines have increased use of resources

As the resources increase, benefits are broadly distributed (green line) beyond the top 20 users (colored lines)

Use of all resources/services, including XSEDE, Blue Waters and NCAR are well characterized Additional Trend Analysis beginning: – Modest increases in parallelism (number of cores used) and job size – Emerging scientific/workflow/data analytic trends? – Industry? – Full costs highly competitive Complement with community engagements in the context of all five strategies: – Foundational research to fully exploit parallelism and concurrency…. – Applications research and development…. – Building, testing, and deploying both sustainable and innovative resources into a collaboration ecosystem…. – Comprehensive education and workforce programs… – Development and evaluation of transformational and grand challenge community programs…

Thank you Irene M. Qualters Acting Division Director, ACI

CREDITS Copyrighted material used under Fair Use. If you are the copyright holder and believe your material has been used unfairly, or if you have any suggestions, feedback, or support, please contact Except where otherwise indicated, permission is granted to copy, distribute, and/or modify all images in this document under the terms of the GNU Free Documentation license, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled “GNU Free Documentation license” at Commons:GNU_Free_Documentation_License The inclusion of a logo does not express or imply the endorsement by NSF of the entities' products, services, or enterprises.