US Secretary of Energy Rick Perry announced a new Department of Energy Request for Proposals (RFP), potentially worth up to $1.8 billion, for the development of at least two new exascale supercomputers—and possibly three—to be deployed at DOE National Laboratories in the 2021-2023 timeframe.
The new supercomputers funded through this RFP (CORAL-2 RFP No. 6400015092) will be follow-on systems to the first US exascale system authorized by Secretary Perry this past June, named Aurora, which is currently under development at Argonne National Laboratory (ANL) and scheduled to come online in 2021.
The RFP announced today also envisions the possibility of upgrades or even a follow-on system to Aurora in 2022-2023, depending on an assessment of needs and opportunities at that time.
|Targeted Exascale HPC systems in the RFP|
|ORNL||System delivered in 2021 and accepted in 2022|
|LLNL||System delivered in 2022 and accepted in 2023|
|ANL||Potential system delivered in 2022 and accepted in 2023|
This RFP calls for a system to be deployed at Oak Ridge National Laboratory in Oak Ridge, Tennessee, and for another to be sited at Lawrence Livermore National Laboratory in Livermore, California. The RFP, the product of a collaboration among the three laboratories known as CORAL—the Collaboration of Oak Ridge, Argonne and Livermore—provides an opportunity for US industry to develop at least two new unique system designs, each with a possible cost range of $400-$600 million, with a possibility of a third system with a similar or wholly different design from each of the previous two. Future amounts are subject to congressional appropriations.
The RFP is being managed for CORAL by UT-Battelle LLC, which operates Oak Ridge National Laboratory on behalf of the Department of Energy.
The new systems will provide 50 to 100 times greater performance than the current fastest US supercomputer. They will enable breakthroughs in both science and industry through modeling and simulation, high-performance data analysis, and artificial intelligence and machine learning applications. Potential examples include:
Identifying next-generation materials
Deciphering high-energy physics data
Accelerating industrial product design and reducing cost-to-market
Evaluating options for nuclear security
CING Funding is not presently available. The Laboratories anticipate funding from DOE Office of Science and the National Nuclear Security Administration (NNSA) to become available in a manner that will allow the selected Offerors to perform continuously until completion.
Funding for the RFP is being provided jointly by the . The Office of Science and NNSA are also partners in the Department’s Exascale Computing Project, which is responsible for research and development efforts to accelerate application development, software technology, and hardware and integration for exascale systems.
ORNL intends to pay for its machine by using third-party Lease-To-Own (LTO) financing. ORNL may choose to exercise the option for third-party leasing in the Draft Lease Agreement.
LLNL intends to pay for its machine using direct funding; however, LLNL reserves the right to use third-party LTO financing for a portion of its machine, depending on annual appropriated funding from Congress.
ANL intends to pay for its machine using third-party LTO financing. ANL’s LTO Sample and the Master Lease Agreement apply only if the successful Offeror can offer mutually acceptable financing options.