The DOE ASCR Leadership Computing Challenge (ALCC) provides allocations of computing time on the nation's most powerful supercomputers for high-risk, high-payoff scientific campaigns aligned with DOE priorities. For the 2026-2027 cycle, available resources include 20 million node-hours on Frontier (exascale, Oak Ridge), 16 million node-hours on Aurora (exascale, Argonne), 1 million node-hours on Polaris (Argonne), and 3.5 million combined CPU and GPU node-hours on Perlmutter (NERSC, with 7,000+ NVIDIA A100 GPUs). ALCC projects span modeling, simulation, artificial intelligence, and data analysis. New for 2026-2027: multi-year proposals of up to three years are accepted, and the process has been simplified to a single proposal without a pre-proposal stage. The estimated commercial value of allocations ranges from $500,000 to $10 million or more per project depending on scale. The 2026-2027 application deadline was January 26, 2026; the program cycles annually with the next call expected November 2026.

The Department of Defense FY2026 Defense University Research Instrumentation Program (DURIP) provides funding for U.S. universities to acquire research equipment and instrumentation in areas important to national defense, including AI and machine learning hardware. The program is administered jointly by the Army Research Office (ARO), Office of Naval Research (ONR), and Air Force Office of Scientific Research (AFOSR), with approximately $34 million available and 95 awards anticipated. DURIP funds the acquisition of specialized computing hardware for AI/ML research (GPU clusters, TPUs, neuromorphic processors), robotics and autonomous systems testbeds, sensor arrays and data collection systems for machine learning training, high-performance computing infrastructure for defense-relevant AI research, and laboratory equipment for human-AI interaction studies. The program specifically supports equipment that enhances research-related education in DoD-priority disciplines. While general-purpose computing is not eligible, computing equipment directly supporting DoD-relevant AI research programs qualifies. No cost sharing is required.