Nvidia Pledges $1B to US Science, Plans Seven DOE Supercomputers
TL;DR
- The $1B blends cash grants and in-kind GPU credits; the undisclosed split complicates budget planning at the seven national lab deployments.
- Jensen Huang received a National Medal of Science at the same summit; TechCrunch reports awardees collectively donated roughly $6B to Trump-related efforts.
- Seven new DoE supercomputers span Argonne and Los Alamos, including Solstice, a 100,000-Blackwell-GPU system co-built with Oracle targeting lower-precision AI workloads.
Nvidia committed $1 billion to US scientific discovery over the next five years, a figure The Register tagged as "about 1/60th of its quarterly profits." The pledge was announced Thursday at an event in Washington, DC and ties into the Trump administration's Genesis Mission, launched late last year to apply AI to research in quantum computing, healthcare and energy security.
Alongside the money, the company said it would supply "no fewer than seven new supercomputers" across the Department of Energy's Argonne, Los Alamos and Lawrence Berkeley National Laboratories. Argonne gets Solstice, a 100,000-Blackwell-GPU machine built with Oracle, aimed at AI workloads that "can get by with 16, eight or even four bits of precision." Los Alamos is lined up for Mission and Vision, both running Nvidia's next-generation Vera Rubin chips over the Quantum-X800 InfiniBand fabric.
The context: DOE's last Nvidia-powered flagships, Summit and Sierra, were commissioned "all the way back in 2018."
Not every new DOE flagship runs on Nvidia silicon. Oak Ridge's Discovery, slated to come online in 2029, is being built by HPE's Cray division around AMD's FP64-optimized MI430X GPUs and Venice CPUs, with peak theoretical performance "somewhere between 3.3 and 8.5 exaFLOPS depending on whether or not Oak Ridge gets a facility power upgrade."
The Washington pledge capped a week in which Jensen Huang collected one of six National Medals handed to tech CEOs at a Trump AI summit; our tracker has logged 199 Nvidia stories in the last 90 days.
What others are reporting
-
NVIDIA Newsroom Read →
First-party press release detailing the three commitment areas: higher-education research, U.S. quantum computing development, and government-focused cloud service providers.
With a $1 billion investment, NVIDIA is putting advanced Super Intelligence in the hands of America's scientists to accelerate breakthroughs in medicine, energy and materials.
-
TechCrunch Read →
Places the Nvidia pledge inside the broader summit: awardees donated roughly $6B to Trump-related efforts, adding a political-economy lens absent from pure-announcement coverage.
Together, the awardees have donated nearly $6 billion to efforts tied to Trump and his administration.
-
Quantum Computing Report Read →
Flags the undisclosed cash-vs-in-kind split as a national lab budgeting risk and details the hybrid GPU-QPU deployment model linking HPC racks with cryogenic quantum hardware.
By linking GPU-accelerated classical supercomputing with physical QPU control stacks, the initiative provides labs with bandwidth to model complex quantum noise channels.
-
The Next Web Read →
Grounds the pledge in Nvidia's 20-year national lab partnership history, citing a Fermilab quantum project that cut accelerator design timelines from five months to three weeks.
Originally reported by theregister.com
Read the original article →Original headline: Nvidia Pledges $1B Over Five Years to US Scientific Computing, Plans Seven DoE Supercomputers