The Office of Advanced Scientific Computing Research (ASCR) of the Office of Science (SC), U.S. Department of Energy (DOE), hereby announces its interest in receiving peer-reviewable Field Work Proposals (FWPs) from integrated teams of scientific researchers, applied mathematicians, computer scientists and computer architects with the goal of standing up an Exascale Co-Design Center. Co-design refers to a computer system design process where scientific problem requirements influence architecture design and technology and constraints inform formulation and design of algorithms and software. To ensure that future architectures are well-suited for DOE target applications and that major DOE scientific problems can take advantage of the emerging computer architectures, major ongoing research and development centers of computational science need to be formally engaged in the hardware, software, numeric methods, algorithms, and applications co-design process that will be responsible for making key tradeoffs in the design of exascale systems. Multi-institutional proposals built upon co-design principles and focused on exploring new scientific problem formulations, algorithms, and execution programming models in close collaboration with the design of exascale computer systems and software are encouraged.
Exascale computer systems are expected to be comprised of as many as a billion cores. Such systems will be capable of 10 billion-way concurrency in simultaneous operations. Industry reports indicate that because of power constraints, data movement, rather than computational operations, will be the limiting factor for exascale systems. Memory per core is expected to decline sharply while the performance of storage systems will continue to lag far behind. Multi-level storage architectures that span multiple types of hardware are anticipated and will require new approaches to run-time data management and analysis.
This Program Announcement invites proposals for integrated co-design centers to address a variety of challenges in creating scientific codes for exascale computing systems, including the design and development of hardware, system software, architectural-aware numeric methods, algorithms and new formulations of the basic application problem. It is anticipated that these co-design centers will need to address: application formulation, advanced programming languages, integrated uncertainty quantification, validation and verification, new mathematics and approaches to implementing multi- physics problems that naturally express parallelism and locality, and data management analysis, including visualization; all in the context of a scientific problem area important to the DOE.