ExaDigiT Digital Twin Framework
Category: Other
Developers: Oak Ridge National Laboratory
Product Description:ExaDigiT is the first comprehensive digital twin framework for modeling liquid-cooled supercomputers, integrating workload, power, and cooling simulations with interactive visualization in AR/VR. It enables holistic system optimization, supporting "what-if" analyses, future system prototyping, and AI-driven control strategies. Developed at ORNL in collaboration with Hewlett Packard Labs, ExaDigiT has been validated on Frontier, the world’s first exascale supercomputer, and the framework’s open-source ecosystem has gained global adoption, with contributions from major supercomputing centers such as CSC (Finland), CINES (France) and Argonne National Laboratory. ExaDigiT’s breakthrough approach connects dynamic power modeling with transient cooling behavior, offering a new way to investigate energy efficiency and optimize HPC operations. It is already influencing next-generation system designs and vendor evaluations, making it an essential tool for the future of sustainable supercomputing.
Developers: Oak Ridge National Laboratory
Product Description:ExaDigiT is the first comprehensive digital twin framework for modeling liquid-cooled supercomputers, integrating workload, power, and cooling simulations with interactive visualization in AR/VR. It enables holistic system optimization, supporting "what-if" analyses, future system prototyping, and AI-driven control strategies. Developed at ORNL in collaboration with Hewlett Packard Labs, ExaDigiT has been validated on Frontier, the world’s first exascale supercomputer, and the framework’s open-source ecosystem has gained global adoption, with contributions from major supercomputing centers such as CSC (Finland), CINES (France) and Argonne National Laboratory. ExaDigiT’s breakthrough approach connects dynamic power modeling with transient cooling behavior, offering a new way to investigate energy efficiency and optimize HPC operations. It is already influencing next-generation system designs and vendor evaluations, making it an essential tool for the future of sustainable supercomputing.
