NASA has officially activated its latest supercomputer, Athena, at the Modular Supercomputing Facility within the Ames Research Center in California. Boasting a performance of 20.13 petaflops, Athena significantly surpasses its predecessor, the Pleiades system, which was decommissioned earlier this month with a capacity of 7.09 petaflops.
According to NASA, Athena is built with 1,024 AMD Epyc Turin nodes and offers 786TB of memory. The system will be accessible to NASA researchers as well as external scientists and researchers supporting NASA projects, who can apply for time on the supercomputer.
“Exploration has always driven NASA to the edge of what’s computationally possible,” said Kevin Murphy, NASA’s Chief Science Data Officer. “With Athena, NASA will expand its efforts to provide tailored computing resources that meet the evolving needs of its missions.”
While Athena is a CPU-based supercomputer, NASA’s overall high-end computing (HEC) fleet includes other powerful systems: Cabeus (20.67 petaflops), Aitken (13.12 petaflops), Electra (8.32 petaflops), Discover (8.1 petaflops), and Endeavour (154.8 teraflops).
However, a 2024 report by the NASA Office of Inspector General (OIG) highlighted challenges faced by the agency’s HEC infrastructure. The report found that NASA’s missions are being hampered by insufficient high-end computing resources, largely due to heavy reliance on CPUs and systems being “oversubscribed and overburdened.” This has resulted in Mission Directorates requesting more computing time than available.
To address these issues, the OIG recommended establishing an executive leadership team to better align NASA’s HEC resources with the agency’s specialized needs.
In March 2025, NASA enhanced its fleet by adding 350 Nvidia GH200 nodes to Cabeus, boosting its total compute power by over 13 petaflops and doubling its peak capability.