NSF High Performance Computing System Acquisition: Towards a Petascale Computing Environment for Science and Engineering
WHAT: Limited Submission Proposal for the National Science Foundation (NSF) High Performance Computing System Acquisition: Towards a Petascale Computing Environment for Science and Engineering.
WHO: Texas A&M University faculty, including TAES, TEES and TTI personnel.
WHY: NSF anticipates making four types of awards: a data-intensive, high-performance computing system ($19.0 million dollars for fouryears); an experimental high-performance computing system of innovative design ($12.0 million dollars for five years); an experimental, high-performance grid test-bed ($10.0 million dollars for four years); and a pool of loosely coupled grid-computing resources ($8.0 million dollars for four years).
WHEN: Internal proposal deadline of July 28, 2008.
HOW: Faculty should submit an internal proposal via email for review by the internal selection committee.
THE FINE PRINT:
The funding agency below has limited the number of solicitations to one proposal per institution. If the number of faculty wishing to submit a proposal exceeds the number allowed by the agency, we will conduct an internal selection process. Below are due dates for the program, including the due date for the internal proposal for review by the internal selection committee, the date for announcement of the internal selection and the due date for submission to the agency.
AGENCY: National Science Foundation (NSF)
AGENCY PROGRAM TITLE: NSF 08-573 High Performance Computing System Acquisition: Towards a Petascale Computing Environment for Science and Engineering
BRIEF PROGRAM DESCRIPTION: The High Performance Computing System Acquisition: Towards a Petascale Computing Environment for Science and Engineering Program is designed for resource provider organizations who are committed to the acquisition and deployment of state-of-the-art HPC systems. This competition emphasizes the provision of systems that deliver high levels of performance for many different types of science and engineering applications. An acquisition may include: computing hardware, local on-line storage, archival storage, wide-area network connection or system software. Conversely, proposals that request support for HPC research will be deemed ineligible and returned without review.
Competitive HPC systems will: expand the range of computationally-challenging science and engineering applications that can be tackled with the TeraGrid HPC portfolio; incorporate reliable, robust system software essential to optimal sustained performance; and provide a high degree of stability and usability.
The system performance on an appropriate set of performance benchmarks will be a factor in the selection of the system(s) to be installed. These benchmarks will be designed to capture the salient attributes of those science and engineering applications placing the most stringent demands on the systems to be provisioned.
The performance requirements and benchmarks for this competition were posted on the NSF web site at here in November 2005. The actual results of any benchmarks used should be submitted in the “Supplementary Documents” section of the proposal.
The submission of benchmark results or estimated benchmark results is required as part of each proposal. Two types of benchmarks are required: projected performance on the standard set of benchmarks posted here and projected performance on benchmarks of the proposer’s own choosing that demonstrate the unique features of the proposed system.
NSF is interested in receiving four types of proposals for the following types of systems: a data-intensive, high-performance computing system that is optimized to support research with very large data-sets or very large input-output requirements that has a total peak computing capacity of the system of at least 200 teraflop/s; an experimental high-performance computing system of innovative design with an architectural design that is outside the mainstream of what is routinely available from computer vendors and that will be an experimental TeraGrid resource used by a smaller number of researchers than is typical for a large TeraGrid resource; an experimental, high-performance grid test-bed which should consist of a small number of high-performance clusters, with heterogeneous architectures, connected by high-bandwidth network links, with the intent of making this resource available to groups of grid researchers who may wish to deploy their own networking, middleware and application software stacks to perform a variety of different types of research in grid computing; and a pool of loosely coupled grid-computing resources which should include a commitment to provide access to a portion of the campus grid that, on average, is at least 100 teraflops of peak computing capacity at the beginning of the period of operational support and that does not decrease over the lifetime of the award.
In addition, the project description must include the following eight sections: system specification; system performance on science and engineering applications; system reliability and usability; implementation, project management and risk mitigation; quality of the physical infrastructure; plan for operations, including and user support and training; integration of research and education; and broadening participation.
The full announcement can be found at the web page of the sponsor here.
AGENCIES ELIGIBLE TO SUBMIT: The High Performance Computing System Acquisition: Towards a Petascale Computing Environment for Science and Engineering Program is open to U. S. institutions of higher education and Federally Funded Research and Development Centers (FFRDC).
AWARD INFORMATION: NSF anticipates making four types of awards: a data-intensive, high-performance computing system ($19.0 million dollars for four years); an experimental high-performance computing system of innovative design ($12.0 million dollars for five years); an experimental, high-performance grid test-bed ($10.0 million dollars for four years); and a pool of loosely coupled grid-computing resources ($8.0 million dollars for four years).
ELIGIBLE PROJECTS COSTS: Indirect costs (F&A) Limitations: NONE.
COST SHARING: NONE.
INTERNAL SELECTION PROCEDURES AND DEADLINES
July 25, 2008: Deadline for an email of intent, including the name of the PI, co-PIs, title of internal proposal and a 1-3 sentence description of the project. Send email of intent.
July 28, 2008: Deadline to obtain approval from your department head and college dean and to submit an internal proposal to the Office of Sponsored Projects. The internal proposal should include:
(1) An executive summary, up to three pages, based on the proposal description as described in the NSF program announcement;
(2) Summary budget;
(3) Project and Management Plans (including team members).
The form for completing the internal proposal is here. The internal proposal must be prepared on the form and sent as one pdf file.
This completed form should be sent electronically for review by the internal selection committee.
*Once your internal proposal has been received with all of the necessary signatures, you will receive an email indicating it will be reviewed by the internal selection committee. If you do not receive the confirmation email, please call 5-1812.*
Please read the RFP carefully for specific requirements of the program here.
Selection of internal proposals will be based on National Science Foundation guidelines. Meeting the needs of the university’s reinvestment plan will also be taken into account.
During the selection process, the internal selection committee may contact departments and colleges for their opinions and commitments. They may also request additional information from PI’s.
August 8, 2008: The Internal Selection committee will notify PIs the results of the internal competition.
November 28, 2008: NSF Deadline for full proposal due by 5 p.m. submitter’s local time.
