Scientific Discovery through Computing (SCiDAC): Partnerships in Basic Energy Sciences

The U.S. Department of Energy’s Office of Science has issued a Funding Opportunity Announcement (FOA) inviting interdisciplinary teams to submit applications for partnerships under the Scientific Discovery through Advanced Computing (SciDAC) program. These partnerships aim to advance scientific discovery in basic energy sciences and computing research by leveraging high-performance computing (HPC) resources. Applications must align with the missions of Basic Energy Sciences (BES) and Advanced Scientific Computing Research (ASCR) and focus on one of three targeted research areas: complex dynamical systems in energy-relevant systems and materials, explainable artificial intelligence for chemical and physical mechanism extraction, or foundation models for chemical and material sciences. The program seeks transformative advancements in algorithms, models, and computational methods that extend current scientific capabilities. The SciDAC program encourages interdisciplinary collaboration between domain scientists (e.g., materials science, physics, chemistry, biosciences) and experts in applied mathematics and computer science. Applicants are expected to collaborate with SciDAC Institutes to ensure effective integration of computational science expertise. Proposed projects must use DOE HPC facilities such as NERSC, ALCF, and OLCF, including exascale systems like Aurora and Frontier. Successful applications will address fundamental research challenges, advance the understanding of complex systems, and promote synergies between computational and domain sciences. The FOA provides a total of $40 million in funding for up to six awards, with individual awards ranging from $1 million to $2.5 million annually for up to four years. Eligible applicants include all domestic institutions except certain nonprofit organizations that engaged in lobbying activities after 1995. Federally affiliated entities like DOE National Laboratories and FFRDCs may participate but must follow specific conditions. Pre-applications are mandatory and must align with one of the three research topics; applications that fail to meet this criterion may be declined without review. Cost-sharing is not required, and projects must focus on basic research with outcomes that are publishable and non-commercial in nature. Applications must demonstrate how the proposed research will exploit HPC resources to achieve scientific breakthroughs, detailing the role of SciDAC Institute collaborators. Teams may include multi-institutional partnerships, but only one application may be submitted per team, with subrecipients identified as needed. The FOA limits each institution to two pre-applications and applications as lead, and no single principal investigator should participate in more than two submissions. Key dates include a pre-application deadline (required before full application submission), with additional details and registration for an informational webinar on January 30, 2025, available through the BES Funding Opportunities webpage. The application process involves detailed narrative and budgetary submissions, data management plans, and a description of partnerships with SciDAC Institutes. DOE emphasizes openness in research and compliance with export controls and data-sharing requirements. Evaluation criteria focus on the scientific and technical merit of the application, the competency of the research team, and the feasibility of the proposed research plan. Reviewers will also consider the integration of computational expertise, the potential for transformative impact, and the alignment with DOE mission priorities. Awardees will be required to participate in annual meetings and demonstrate progress through regular reports. Funding decisions will prioritize innovative approaches that address fundamental research challenges and leverage DOE’s HPC capabilities effectively.