Skip to main content
v2026.11,986 entries · CC-BY 4.0

Editorial · CASRAI · Funding lifecycle and financial vocabulary

Schmidt Sciences awards $334 million for NCSA research supercomputer

A $334 million Schmidt Sciences grant will fund NCSA’s Garnet supercomputer, with launch expected in mid-2027.

Published 17 Nov 2026· 5 minute read

Ask CASRAI · free to try

Ask about this story

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

CASRAI is the reference for research administration — bookmark it for the next question.

The National Center for Supercomputing Applications announced a $334 million Schmidt Sciences research grant on November 16, 2026, to acquire and operate Garnet, a supercomputer expected to launch in mid-2027. The University of Illinois Urbana-Champaign system joins a wider expansion of institution-owned scientific computing, with parallel announcements from Berkeley and Johns Hopkins offering researchers new capacity and support.

A $334 million grant funds Garnet at Illinois

In its November 16 announcement, NCSA said Garnet would operate within the Schmidt Science Computing Network, or SSCN. The system will be housed at the National Petascale Computing Facility on the Illinois campus and will initially use NVIDIA Blackwell GPUs. NCSA describes the design as capable of incorporating later hardware and emerging GPU technology.

Tim Boerner, NCSA’s associate director of computing infrastructure, will lead the project. Director R. Srikant will serve as chief scientific adviser. NCSA identifies artificial intelligence, astrophysics, biological sciences and climate sciences among the research areas the resource will support. It says Illinois researchers will receive reserved campus allocations.

The network combines hardware, software and scientific support

Schmidt Sciences describes SSCN as a distributed network whose participating universities and research institutions own, operate and maintain their systems. Its grants support GPU and CPU computing, storage and networking, alongside research software and technical expertise. Institutions also contribute computing capability to the network.

The funder frames the programme as a way to build lasting scientific capability. Scientific software, operational knowledge and research support are part of that model, rather than incidental services attached to a machine. Its stated purpose is to complement government-supported and university infrastructure while preserving scientific independence.

For research offices, this is a research infrastructure funding development: an external grant establishes a resource that the host institution will operate for scientific work. The implications extend to the teams that maintain systems and help researchers use them, including research software engineers.

UC Berkeley and Berkeley Lab announced on November 16 that a Schmidt Sciences grant would establish the Berkeley Science AI Platform. It will add computing and storage capacity to Doudna, the National Energy Research Scientific Computing Center’s new flagship system, with deployment described for 2027. Dedicated capacity is intended for Schmidt Sciences awardees and Berkeley researchers.

The Berkeley announcement connects that resource to cosmology, genome engineering, earth science, materials science and computational biology. It identifies NERSC as a Department of Energy scientific user facility and describes the platform as a collaboration linking academic, philanthropic and public institutions.

Johns Hopkins separately announced a gift of advanced computing resources to its Whiting School of Engineering. The university says the gift will provide specialised computing and expanded data capacity for its Data Science and AI Institute. That announcement places the new resources within the same distributed network and links them to research across medicine, public health and other sciences.

Johns Hopkins also identifies an earlier relationship with the funder: in 2022 it joined a multi-university initiative supporting scientific software engineers. The new computing gift builds on that approach by adding infrastructure to the people and software needed for research.

Scientific leaders emphasise access and judgment

In NCSA’s announcement, Srikant said the centre’s experience operating national-scale computing infrastructure would support research across disciplines. Boerner described Garnet as a central computing resource for the network, with the scientific capabilities delivered to researchers as the project’s focus.

Berkeley Lab director Kathy Yelick linked the Berkeley platform to broader access to advanced computing for machine-learning applications and the Department of Energy’s Genesis Mission. Jennifer Doudna, quoted in Berkeley’s announcement, emphasised that greater computing power still requires scientists to choose which possibilities deserve experimental attention. At Johns Hopkins, institute director Mark Dredze connected the new capacity to pursuing scientific questions across disciplines.

What the expansion means for research teams

Independent reporting has highlighted different parts of the expansion. TechTimes’ November 16 analysis focused on Garnet and the importance of matching computing hardware to scientific workloads. SciBeat’s November 17 report examined the Berkeley platform and its intended research capacity. Both developments concern infrastructure for doing science, including the support and allocation arrangements that determine how teams can use it.

The practical implication for university research offices is to connect prospective users with their institution’s computing staff as allocation routes emerge. Planning a scientific workload involves software, data and support requirements as well as access to processors. CASRAI’s background guide on requesting GPUs in a Slurm job explains one common way researchers specify computing resources; actual submission procedures will depend on each system.

Deployment and allocation processes come next

NCSA expects Garnet to launch in mid-2027, while Berkeley describes deployment of its platform in 2027. Schmidt Sciences says access to SSCN will use defined allocation processes managed by participating institutions and selected funder programmes. It will publish specific access information and opportunities as systems and allocation pathways become available. Those institutional processes will turn the new funding into computing time for research teams.

Sources

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →