Nvidia has partnered with Oracle and the US Department of Energy to build two AI supercomputers at Argonne National Laboratory in Illinois.

Named Equinox and Solstice, both systems will be powered by Nvidia Blackwell GPUs and support the development and training of frontier models for use by the scientific community.

Featuring a “record-breaking” 100,000 Blackwell GPUs, Solstice will be the largest AI supercomputer owned by the DOE and will be built with the DOE’s new public-private partnership model. Equinox will consist of 10,000 Blackwell GPUs and is expected to come online in the first half of 2026.

Both systems will be interconnected by Nvidia networking technology – although the company did not specify if this would be its NVLink offering or Spectrum-X Ethernet platform. When both systems are online, they will deliver a combined 2,200 exaflops of AI performance.

Argonne also said it was planning to launch three additional Nvidia-based systems, Tara, Minerva, and Janus, to further expand access to AI compute for researchers across the US. No further information about which Nvidia platform will power those three systems has been disclosed.

“AI is the most powerful technology of our time, and science is its greatest frontier,” said Jensen Huang, founder and CEO of Nvidia. “Together with Oracle, we’re building the Department of Energy’s largest supercomputer that will serve as America’s engine for discovery, giving researchers access to the most advanced AI infrastructure to drive progress across fields ranging from healthcare research to materials science.”

US Secretary of Energy, Chris Wright, added: “Winning the AI race requires new and creative partnerships that will bring together the brightest minds and industries American technology and science has to offer. The two Argonne systems and the collaboration between the Department of Energy, Nvidia, and Oracle represent a new common-sense approach to computing partnerships. These systems will be a powerhouse for scientific and technological innovation. Thanks to President Trump, we’re bringing new computing capacity online faster than ever before and turning shared innovation into national strength.”

LANL and Lilly also set to deploy Nvidia-based systems

The two new DOE supercomputers were not the only large-scale compute deployments announced by Nvidia at its GTC event in Washington, DC, this week.

US pharmaceutical company Lilly will deploy the world’s first Nvidia DGX SuperPOD powered by 1,016 Nvidia Blackwell Ultra GPUs. The SuperPOD will be used to train biomedical foundation and frontier models, reducing drug discovery timelines and enabling accelerated breakthroughs in genomics, personalized medicine, and molecular design.

Elsewhere, Los Alamos National Laboratory in New Mexico announced that it has selected Nvidia’s Vera Rubin platform and Nvidia Quantum-X8000 InfiniBand networking fabric for its forthcoming Mission and Vision AI supercomputers.

Mission will be the fifth Advanced Technology System (ATS5) in the National Nuclear Security Administration’s Advanced Simulation and Computing program. Designed to run classified applications, the system is expected to be available from 2027. Vision, meanwhile, will run unclassified workloads supporting AI and open science research, building on the achievements of LANL’s Venado supercomputer.

“Our integration of the Nvidia Vera Rubin platform and Quantum X800 InfiniBand fabric represents a transformative advancement of our lab — harnessing this level of computational performance is essential to tackling some of the most complex scientific and national security challenges,” said Thom Mason, director of Los Alamos National Laboratory. “Our work with NVIDIA helps us remain at the forefront of innovation, driving discoveries to strengthen the resilience of our critical infrastructure.”