Novel Technologies for High Performance Computing
The Intelligence Advanced Research Projects Activity (IARPA) often selects its research efforts through the Broad Agency Announcement (BAA) process. This request for information (RFI) is intended to provide information relevant to a possible future IARPA program, so that feedback from potential participants can be considered prior to the issuance of a BAA. Respondents are invited to provide comments on the content of this announcement to include suggestions for improving the scope of a possible solicitation to ensure that every effort is made to adequately address the scientific and technical challenges described below. Responses to this request may be used to support development of, and subsequently be incorporated within, a future IARPA Program BAA and therefore must be available for unrestricted public distribution. The following sections of this announcement contain details of the scope of technical efforts of interest, along with instructions for the submission of responses.
Background and Scope
There have always been computational challenges that demand more computing power than is available, and this tension has helped drive decades of dramatic increases in high performance computing (HPC) worldwide.1 The top performing system (as of November 2012) delivers 17.6 PFLOP/s on a well-known benchmark2 while consuming 8.2 MW of electrical power. The U.S. Department of Energy's Exascale Computing Initiative aims to achieve3 1 EFLOP/s using ~20 MW of power within a decade. All of today's leading HPC systems use similar CMOS technologies in a variety of architectures; their continued adequacy for ever more demanding applications hinges on the annual performance improvements that have heretofore been a hallmark of the semiconductor industry.
In this RFI we seek information about novel technologies that have the potential to enable new levels of computational performance with dramatically lower power, space and cooling requirements than the HPC systems of today. Importantly, we also seek to broaden the definition of high performance computing beyond today's commonplace floating point benchmarks, which reflect HPC's origins in the modeling and analysis of physical systems. While these benchmarks have been invaluable in providing the metrics that have driven HPC research and development, they have also constrained the technology and architecture options for HPC system designers. The HPC benchmarking community has already started to move beyond the traditional floating point benchmarks with new benchmarks focused on data intensive analysis of large graphs4 and on power efficiency.5
A variety of alternative technologies with the potential to support high performance computation - in the broadest sense - have appeared in the research literature, but none have made the technical advances needed to compete with the mature HPC industry. Some technologies have the potential to do well on today's benchmarks but lack the maturity to do so (e.g., superconducting computation); others have the potential to achieve high performance on important computational challenges but are highly unlikely to do well on today's benchmarks (e.g., quantum computation, molecular/DNA computation, neural computation, optical computation).
With this RFI we seek responses that illuminate the breadth of technologies that offer the potential to build high performance computation systems to address hard computational challenges in the future, particularly where those challenges are not necessarily represented by today's HPC benchmarks. Some examples of technologies are given above, but these are not intended to be comprehensive. We are not interested in responses that reflect current conventional HPC technologies (e.g., graphical processing units).
Each response should be focused on a specific technology, and provide:
- A comprehensive description of the technical area,
- An assessment of the maturity level, technical challenges, technical community size and distribution across academia, industry and geography, location of prominent researchers, level of research investment today (and sources of that investment: government/industry/other),
- A list of hard and important computational challenges against which the technology is likely to do well, with suggested benchmarks and metrics,
- The model of computation best supported by the technology (particularly where a non-standard model is envisioned),
- Technical areas which will benefit if the computational challenges are addressed, and whether this may lead to a potential commercial market,
- R&D timeline(s) to demonstrate success, with level(s) of research investments and key milestones and metrics.
Responses to this RFI will be used to decide whether to hold a focused workshop on novel technologies for high performance computation in May/June 2013.
Preparation Instructions to Respondents
IARPA solicits respondents to submit ideas related to this topic for use by the Government in formulating a potential program. IARPA requests that submittals briefly and clearly describe the potential approach or concept, outline critical technical issues/obstacles, describe how the approach may address those issues/obstacles and comment on the expected performance and robustness of the proposed approach. If appropriate, respondents may also choose to provide a non-proprietary rough order of magnitude (ROM) regarding what such approaches might require in terms of funding and other resources for one or more years. This announcement contains all of the information required to submit a response. No additional forms, kits, or other materials are needed.
IARPA appreciates responses from all capable and qualified sources from within and outside of the US. Responses from teams with complementary areas of expertise are encouraged. Responses have the following formatting requirements:
- A one page cover sheet that identifies the title, organization(s), respondent's technical and administrative points of contact - including names, addresses, phone and fax numbers, and email addresses of all co-authors, and clearly indicates its association with RFI-13-02.
- A substantive, focused, one-half page executive summary;
- A description (limited to 5 pages in minimum 12 point Times New Roman font, appropriate for single-sided, single-spaced 8.5 by 11 inch paper, with 1-inch margins) of the technical challenges and suggested approach(es);
- A list of citations (any significant claims or reports of success must be accompanied by citations, and reference material MUST be attached);
- Optionally, a single overview briefing chart graphically depicting the key ideas.
Disclaimers and Important Notes
This is an RFI issued solely for information and new program planning purposes and does not constitute a solicitation. Respondents are advised that IARPA is under no obligation to acknowledge receipt of the information received, or provide feedback to respondents with respect to any information submitted under this RFI.
Responses to this notice are not offers and cannot be accepted by the Government to form a binding contract. Respondents are solely responsible for all expenses associated with responding to this RFI. It is the respondents' responsibility to ensure that the submitted material has been approved for public release by the organization that funded whatever research is referred to in their response.
The Government does not intend to award a contract on the basis of this RFI or to otherwise pay for the information solicited, nor is the Government obligated to issue a solicitation based on responses received. Neither proprietary nor classified concepts or information should be included in the submittal. Input on technical aspects of the responses may be solicited by IARPA from non-Government consultants/experts who are bound by appropriate non-disclosure requirements.
- See, for example http://www.top500.org which publishes semiannual updates to a listing of the top 500 HPC systems in the world, ranked according to their measured performance on the LINPACK benchmark.
- HPC system performance is reported as how many floating point operations they execute per second on benchmark tests. Today's leading systems have ratings that use the metric peta prefix, e.g., 1 PFLOP/second (or "1 PFLOP/s") refers to 1015 floating point operations per second. Similarly, "tera" (1012) is written TFLOP/s, and "exa" (1018) is written EFLOP/s.
For information contact:firstname.lastname@example.org
IARPA-RFI-13-02 CLOSEDPosted Date: March 8, 2013
Responses Due: April 5, 2013