Documentation on High Performance Computing

The following documents were kindly supplied by Richard Loft (SCD) to help motivate further discussion on the question of NCAR's response to NSF's $30M RFP (NSF05-625) for a high-performance computing facility. Dr. Loft would appreciate input on the subject from the NSA membership.  You may send your comments to loft@ucar.edu

"High Performance Computing System Acquisition: Towards a Petascale Computing Environment for Science and Engineering"

Program Solicitation NSF 05-625
National Science Foundation, 2005

http://www.nsf.gov/funding/pgm_summ.jsp?pims_id=13649&org=OCI&from=home
http://www.isse.ucar.edu/HP_rick/NSA/nsf05625.pdf

The purpose of this solicitation is to generate proposals from Resource Provider organizations who are committed to the acquisition and deployment of balanced HPC systems that will contribute to the development of the HPC environment described in the Introduction. In future competitions, these systems will be upgraded or complemented by the acquisition of additional HPC systems that may be optimized for particular classes of science and engineering research problems. However, this competition emphasizes the provision of one or two well-balanced systems that deliver high levels of performance for many different types of science and engineering applications.

"Getting up to speed: The future of supercomputing"


Committee on the Future of Supercomputing, 2005
National Research Council

http://www.isse.ucar.edu/HP_rick/NSA/FOSCfinalreport.pdf

The Computer Science and Telecommunications Board (STB) of the National Research Council (NRC) convened the Committee on the Future of Supercomputing to assess prospects for supercomputing technology research and development in support of U.S. needs, to examine key elements of context -- the history of supercomputing, the erosion of
research investment, the changing nature of the problems demanding supercomputing, and the needs of government agencies for supercomputing capabilities -- and to assess opportunities for progress.

"ESTABLISHING A PETASCALE COLLABORATORY FOR THE GEOSCIENCES: SCIENTIFIC FRONTIERS"

Report of the Ad Hoc Committee and Technical Working Group for a Petascale Collaboratory for the Geosciences, 2005.

http://www.isse.ucar.edu/HP_rick/NSA/petascale_science.pdf

This report is the outcome of research and deliberations of an ad hoc committee of scientists working on behalf of the atmospheric, solid Earth, ocean, and space science communities, which was formed to address this gap between the scientific requirements for, and the availability of, high-end computational resources. The committee
was formed with NSF's encouragement. The report presents an overview of the scientific frontiers that would be opened by a national investment in leadership-class computing systems dedicated to geosciences research. The companion report (Technical Working Group and Ad Hoc Committee for a Petascale Collaboratory for the Geosciences, 2005) describes a technical and budgetary prospectus for one possible implementation scenario for a Petascale Collaboratory for the Geosciences.


"ESTABLISHING A PETASCALE COLLABORATORY FOR THE GEOSCIENCES: TECHNICAL AND BUDGETARY PROSPECTUS"

Report of the Technical Working Group and Ad Hoc Committee for a Petascale Collaboratory for the Geosciences, 2005.

http://www.isse.ucar.edu/HP_rick/NSA/petascale_tech.pdf

The technical feasibility of the PCG's [Petascale Collaboraory for the Geosciences] scientific vision is determined by the architectural trends in the supercomputing industry, the current and projected computational requirements of the applications, and the overall cost of the project. This document evaluates each of these aspects and discusses, in general terms, how best to construct the PCG within a five-year time frame, starting in 2007.


"Petascale Computational Systems:  Balanced CyberInfrastructure in a Data-Centric World"

Gordon Bell (1), Jim Gray (1) and Alex Szalay (2)
(1) Microsoft Research
(2) The Johns Hopkins UniversitySeptember 2005

http://www.isse.ucar.edu/HP_rick/NSA/Petascale_systems.pdf

Computational science is changing to be data intensive. NSF should support balanced systems, not just CPU farms but also petascale IO and networking. NSF should allocate resources to support a balanced Tier-1 through Tier-3 national cyber-infrastructure.