O C I NSF s Office of Cyberinfrastructure Kevin

O C I NSF’s Office of Cyberinfrastructure Kevin Thompson Program Director National Science Foundation Office of Cyberinfrastructure kthompso@nsf. gov (many of the slides Are courtesy of Dr. Atkins) 1

NSF Blue Ribbon Advisory Panel on Cyberinfrastructure “a new age has dawned in scientific and engineering research, pushed by continuing progress in computing, information, and communication technology, and pulled by the expanding complexity, scope, and scale of today’s challenges. The capacity of this technology has crossed thresholds that now make possible a comprehensive “cyberinfrastructure” on which to build new types of scientific and engineering knowledge environments and organizations and to pursue research in new ways and with increased efficacy. ” vhttp: //www. nsf. gov/od/oci/reports/toc. jsp O C I Daniel E. Atkins, Chair University of Michigan Kelvin K. Droegemeier University of Oklahoma Stuart I. Feldman IBM Hector Garcia-Molina Stanford University Michael L. Klein University of Pennsylvania David G. Messerschmitt University of California at Berkeley Paul Messina California Institute of Technology Jeremiah P. Ostriker Princeton University Margaret H. Wright New York University 2

O Some Science Drivers I v Inherent complexity and multi-scale nature of todays v v v C frontier science challenges. Requirement for multi-disciplinary, multi-investigator, multi-institutional approach (often international). High data intensity from simulations, digital instruments, sensor nets, observatories. Increased value of data and demand for data curation & preservation of access. Exploiting infrastructure sharing to achieve better stewardship of research funding. Strategic need for engaging more students in high quality, authentic science and engineering education. 3

Cyber. Infrastructure is about Connectedness between People Ideas Systems Orgs 4 O C I

Investing Within the Framework of the NSF CI Vision O C I Advances in components of CI-systems for S&E R&E Complex, multi-scale, multidisciplinary S&E research challenges Call for Action Blue Ribbon Panel reports plus 30+ disciplinary or interdisciplinary community workshops on CI Framework for Action NSB and NSF internal working groups 5

O C I 6

CI Vision for 21 st Century Discovery O C I High Performance Computing 7

O Drivers for HPC Strategy I v Modeling, simulation, prediction – – C complex systems width: multi-disciplinary, more systemic depth: multi-scale community codes (complex collaborations) v Extraction of knowledge/discovery from massive collections of heterogeneous data v More powerful visualization and interaction capabilities 8

HPC Multi Track Strategy O C I Track 1: One solicitation funded over 4 years: $200 M acquisition + additional O&M cost. Track 2: Four solicitations over FY 06 -09: $30 M/yr acquisition + additional O&M cost. 11

Tera. Grid: 11 Resource Providers, One Facility O I Grid Infrastructure Group (UChicago) UW PSC UC/ANL NCAR PU NCSA IU Caltech UNC/RENCI ORNL Tennessee USC/ISI SDSC TACC C LSU Resource Provider (RP) Software Integration Partner Network Hub 14

O The Tera. Grid Offers C I v A portfolio of high performance computing and visualization architectures v Common user environments v Pooled community support expertise v Targeted consulting services (ASTA) v Science gateways to simplify access, support collaboration, and integration of research and education. v Knowledge management & collaboration infrastructure 15

Science Gateways (Virtual Organizations) O C I v. Built to serve communities of practice by bringing together a variety of resources in a customized portal v. Examples include: • Nano. Hub (models, analysis tools, education tools for nano-science and nano-engineering) • NEES (A distributed earthquake engineering collaboratory) • LEAD (A gateway to support on-demand modeling and analysis of tornados and other strong storms) • SCEC Earthworks Project (Earthquake hazard assessment from first principles) • NVO (National Virtual astronomy Observatory - increasingly used vehicle for astronomical study) vhttp: //www. teragrid. org/programs/sci_gateways/

Projected Tera. Grid Performance Growth (with. Track 1) O C I • 14 PF/s • >1000 TB largest memory 17

TACC Track-2 a v. Compute power - 504 Teraflops aggregate peak – 3, 936 Sun four-socket, quad-core nodes – 15, 744 AMD Opteron “Barcelona” processors l Quad-core, four flops/cycle (dual pipelines) v. Memory – 2 GB/core, 32 GB/node, 125 TB total – 132 GB/s aggregate bandwidth v. Disk subsystem – 72 Sun x 4500 “Thumper” I/O servers, 24 TB each – 1. 7 Petabyte total storage v. System power: 3. 0 MW total v. System: 2. 4 MW – ~90 racks, in 6 row arrangement – ~100 in-row cooling units – ~4000 sq. ft. total footprint v. Infiniband interconnect – Full non-blocking 7 -stage Clos fabric – Low latency (~2 msec), high-bandwidth (~950 MB/s) 18 O C I

Tera. Grid HPC Usage by Discipline 4/01/06 – 3/31/07 GEO 6% CISE 3% SBE DMS 0. 2% 0% C I Staff 0% BIO 23% DMR 9% ENG 10% O l CHE 19% PHY 15% AST 15% Courtesy of John Towns

CI Vision for 21 st Century Discovery O C I High Data & Performance Visualization Computing /Interaction 20

Drivers for Data & Interaction Strategy O C I v v v Increasing Scale and Heterogeneity of Data Integration, federation, interoperability IP issues, sharing, openness Curation, quality control Long-term stewardship/preservation (including appraisal) v Operational sustainability v Appropriate workflow, visualization environments 21

Data. Net Partners: Three Primary Goals O C I v Building information integration capability on the foundation of a reliable data preservation network. v Achieving long-term preservation/access capability in an environment of rapid technology advances. v Achieving economic and technological sustainability. 23

New Types of Organizations Envisioned O C I v Integrate library and archival sciences, cyberinfrastructure, computer and information sciences, and domain science expertise. v Provide reliable digital preservation, access, integration, and analysis capabilities for science and/or engineering data over decades-long timeline. v Continuously anticipate and adapt to changes in technologies and in user needs and expectations v Engage at the frontiers of computer and information science and cyberinfrastructure with research and development to drive the leading edge forward; and v Serve as component elements of an interoperable data preservation and access network. 24

Earth Observing Systems NEON WATERS OOI ORION CUAHSI GEON Tools Interoperability Data and Computation DATANET Foundation O C I

CI Vision for 21 st Century Discovery NVO Virtual Organizations for Distributed Communities O I CMS LEAD High Data & Performance Visualization Computing /Interaction ATLAS i. VDg. L Nano. Hub NEES C 26

O Drivers for Virtual Organization Strategy v Research community driven demand for distributed multi-disciplinary, multi-institutional, multiobservational, multi-facility science projects. v Need for new approaches to broadening participation in both research/discovery and passion-for-science building, inquiry-based, learning/education. 27 C I

Benefits of Virtual Organizations O v Possibilities for “better than being there” organizational forms: – decreased time to discovery; – decreased time from discovery; – increased intellectual cross-section and transformational results; – enhanced stewardship and Ro. I for research infrastructure investments; – multi-use: discovery, learning, rapid-response, . . v A key to transforming the What, How, and Who participates. v A key to economic leadership in a global knowledgebased flat world. 28 C I

Virtual Organizations offer additional modes of interaction between People, Information, and Time Same Different Facilities Geographic Place Different Same (synchronous) (asynchronous) ST-SP P: Physical mtgs I: Print-on-paper books, journals F: Physical labs, studios, shops DT-SP P: Shared notebook I: Library reserves F: Time-shared physical labs, . . . ST-DP P: AV conference I: Web search F: Online instruments DT-DP P: Email I: Knowbots F: Autonomous observatories P: people, I: information, F: facilities, instruments 29 O C I

VO-Global: International R&E Networking O C I 30

NSF’s International Research Connections (IRNC) Program v Goal - enable international science and engineering research collaborations and activities involving U. S. scientists, engineers, and educators § 5 year program (2004 -2009) and 5 main awards for US $25 M total: §US-China and other partners (GLORIAD) §US-Japan and TEIN 2 partners (Trans. PAC 2) §US-Europe (Translight/Starlight) §US-Latin America (WHREN) §US-Australia (Trans. Light/Paacific. Wave) 31 O C I

O But What About the Social Architecture? C I v Many “collaboratories” have been less effective than hoped or even outright failures. Reasons are often more social than technical. v Need more principled understanding of the analysis and design of virtual organizations from a combined socio-technical perspective is critical to achieving the flexibility and agility to respond to new and emerging challenges in an increasingly competitive knowledge-based economy. 32

O VO Challenges C I v Need interdisciplinary experimental research involving practicing science and engineering communities to create more systematic knowledge about the intertwined social and technical issues of effective virtual organizations and how they can act as a collective force to change not only how we practice research but what we produce from it. v Leaders of science & engineering virtual organizations need help in understanding social architecture design principles. v Need better understanding the forms and attributes of new organizational forms made possible by CI. 33

CI Vision for 21 st Century Discovery O C I Virtual Organizations for Distributed Communities High Data & Performance Visualization/I Computing nteraction Learning & Work Force Needs & Opportunities 34

Drivers for LWD Strategy O C I v Education and workforce development to create and use CI for S&E research and education. v More effective learning through the application of cyberinfrastructure. v Exploiting the new opportunities that cyberinfrastructure brings for broadening participation by people who, because of physical capabilities, location, or history, have been excluded from the frontiers of scientific and engineering research and education. v Explore CI support for integrated research and education. 35

Possible Next Frontiers for CI-enabled Learning and Workforce Development O C I v Develop authentic cyberlearning opportunities through the creation of virtual environments that provide laboratory -based research experiences to enhance formal and informal education. v Enhance teacher training for and with cyberinfrastructure tools and resources to prepare a 21 st century teaching force for a 21 st century workforce. v Use cyber tools and resources to collect and analyze firsttime data pertaining to individual and organizational learning to improve our understanding of human cognition and meta-cognition, action and interaction. v Investigate the use of handheld mobile devices as platforms for cyberlearning delivery and discovery. 36

OCI Funding Opportunities O C I v High Performance Computing – Tera. Grid III – informational meeting June 25 at NSF – Track II 2009 – stay tuned – Peta. Apps – stay tuned v Data – Data. Net – Sustainable Digital Data Preservation and Access Network Partners, 07 -601, next prelim proposal due date October 6, 2008 v Virtual Organizations – Virtual Organizations as Sociotechnical Systems (VOSS), NSF 08 -550, deadline was June 2, 2008 v Other programs – Strategic Technologies for Cyberinfrastructure (STCI) l next target date August 14, 2008 – International Reserach Network Connections (IRNC) program l solicitation expected early 2009 37

O C I Thank You 38