How Much Do Supercomputers Cost
Supercomputer prices range from a few hundred thousand dollars for small clusters to over $600 million for the world's largest systems, with most large-scale deployments falling between $50 million and $300 million depending on performance, scale, and infrastructure requirements. The cost depends on processor count, interconnect technology, cooling systems, power infrastructure, and software stack, with specialized components like high-bandwidth memory and custom networking often driving expenses higher than standard data center hardware.
Entry-level research clusters and departmental systems typically cost between $500,000 and $5 million, while mid-range systems used by universities and national labs generally run from $10 million to $100 million. Enterprise and government installations targeting top performance tiers invest $150 million to over $600 million, with the Frontier system at Oak Ridge National Laboratory and the El Capitan system at Lawrence Livermore National Laboratory representing the higher end of current public deployment figures.
Top Supercomputer Prices and Rankings
Current Top Systems and Their Costs
The Frontier exascale system at Oak Ridge National Laboratory, delivered by HPE Cray in 2022, was reported to cost approximately $600 million and held the top position on the TOP500 list for multiple ranking cycles, while El Capitan at Lawrence Livermore, also built by HPE Cray, is estimated at a similar scale with a target performance exceeding two exaflops.
The LUMI system in Finland, built by HPE Cray and operational since 2023, cost around $150 million and ranks among the fastest systems in Europe, while the Leonardo system in Italy, supplied by Atos, cost an estimated $250 million and uses Intel Xeon processors and Nvidia GPUs to deliver multi-petaflop performance. Other notable systems such as Fugaku, Perlmutter, and Summit illustrate a wide cost range driven by architecture choices, vendor contracts, and site-specific infrastructure needs.
Factors That Influence Supercomputer Pricing
Key cost drivers include the number and type of compute nodes, interconnect bandwidth, storage capacity, power and cooling infrastructure, and software development, with custom designs and long-term support contracts adding significant expense beyond raw hardware.
Deployment costs also include facility construction, site preparation, and ongoing operations, as supercomputers require specialized data centers with robust electrical grids, redundant cooling, and high-capacity networking, making total ownership cost substantially higher than the initial hardware purchase price.
Supercomputer Market and Vendor Landscape
Major Vendors and Contracts
HPE Cray, Lenovo, Dell Technologies, and Atos are leading vendors that build large-scale supercomputers for governments, research institutions, and enterprises, with contracts often running into hundreds of millions of dollars and spanning multiple years for hardware, integration, and support.
In 2024, HPE secured a contract to build a next-generation supercomputer for the U.S. Department of Energy, continuing a trend of major federal investments in high-performance computing, while companies like Nvidia supply accelerators and interconnect technologies that form a significant portion of system cost and performance.
How Companies and Organizations Fund Supercomputers
National governments, defense agencies, and research institutions fund most large supercomputers through direct capital budgets and multi-year programs, with the U.S. Department of Energy, European agencies, and Asian governments representing the largest buyers.
Private companies increasingly access supercomputing power through cloud-based high-performance computing services and partnerships rather than owning systems outright, allowing them to use top-tier clusters for AI training, drug discovery, and complex simulations without multi-hundred-million-dollar capital outlays.