Dario Gil at GTC: DOE’s Genesis Mission Moving from Vision to Execution
At Nvidia’s GTC conference last week, a fireside chat between Nvidia’s Ian Buck and U.S. Department of Energy Under Secretary for Science Dario Gil focused on a familiar theme for the HPC community: the convergence of advanced computing and scientific discovery. The conversation highlighted the next phase of the DOE’s Genesis mission, as the department begins turning its AI for science agenda into teams, systems and funded research programs.
The Genesis Mission, launched last November through an executive order, is a national initiative aimed at transforming the way science is conducted in the U.S.
“What we seek to do is to revolutionize the practice of science and engineering with advanced computing. And we seek to combine high performance computing, AI supercomputers, and quantum computers, and bring them into a unified platform to accelerate science and engineering. It’s an incredibly exciting time,” Gil told Buck.
Nvidia VP of Hyperscale and HPC Ian Buck and DOE Under Secretary Dario Gil at a fireside chat at Nvidia GTC last week
Gil, who is leading the initiative, described Genesis as both an infrastructure buildout and a coordinated effort to address a defined set of national challenges. The DOE has already identified 26 challenges spanning energy, discovery science and national security. Examples include accelerating the path to commercial fusion, reducing the cost and timeline of nuclear power, and advancing biotechnology and materials science.
Buck described those challenges in the context of the DOE’s existing lab infrastructure, where many of these problems are already being explored at scale: “I recently visited Argonne National Labs and got to see the APS and the light sources and how it all connects to some of the world’s largest supercomputers. And now, with the advent of AI, we have both simulation and AI that can help solve some of these problems,” Buck said. “You have some of the most passionate scientists working at these laboratories, and they’re working on some of the grand challenges that actually face our nation.”
Rather than treating AI as a standalone domain, the Genesis Mission places it directly inside scientific workflows. The goal is not simply to scale models, but to change how experiments are designed, simulations are run and data is interpreted. Gil gave the example of fusion research. After decades of experimental work and the development of highly accurate simulation codes, researchers are now using AI surrogate models to dramatically increase simulation speed. Gil said these approaches can accelerate certain workflows by orders of magnitude, speeding up the time between simulation and experiment.
The Genesis mission is built around this pattern, where AI systems are layered on top of long-standing investments in data, instrumentation and simulation. Gil described the history behind protein structure prediction as another case study. Long before systems like AlphaFold, national lab infrastructure and academic research produced the foundational datasets that made large-scale prediction possible. The implication for Genesis is that similar opportunities exist across other domains, from high energy physics to materials discovery.
The goal is to replicate that trajectory across dozens of fields. Gil suggested that success for Genesis would mean producing not just one breakthrough, but many. In practical terms, that could translate into dozens of AI-enabled advances emerging from existing scientific infrastructure over the next several years.
“What about everywhere else, where we have all of these data sets, all of these exquisite scientific instruments, all of these communities around that? And how can we mobilize them to have not only one story of AlphaFold, but in a few years from now, we would have 50, 100, 200 stories, one after the other, field by field,” Gil said. “We’ll say, this is what it means for microscopy, and this is what it means for high energy physics, and this is what it means for cosmology, and this is what it means for fusion, and so on. And we are going to achieve that.”
To support that goal, the DOE is pairing its traditional role in large-scale facilities with a new model for collaboration. Genesis is structured to encourage collaboration across national labs, industry and universities, particularly as projects move into later stages. Nvidia is one of the early partners in this effort. The company is working with the DOE and other collaborators to deploy new AI-focused systems, including large GPU clusters at Argonne National Laboratory. These systems are designed to provide the computational backbone for Genesis workloads and to make advanced AI infrastructure accessible to the scientific community.
Gil emphasized that speed is a priority. A week ago, the DOE announced an initial $293 million in funding to support the first wave of projects. Teams will begin with smaller, exploratory efforts before scaling into larger, multi-institution programs. The timeline is aggressive, with early teams expected to be operational this summer.
Beyond infrastructure and funding, Gil also highlighted workforce development as a central component. Genesis Mission includes plans to train tens of thousands of scientists and engineers in AI-enabled research methods. Physicists, chemists and biologists will be able to incorporate AI into their existing disciplines, augmenting domain expertise instead of replacing it.
In its effort to build a unified scientific computing platform, the Genesis Mission also includes quantum computing as an emerging component. The long-term vision is for tightly integrated systems that combine classical HPC, AI accelerators and quantum processors. These hybrid architectures are expected to play a role in areas such as chemistry and materials science. Gil mentioned how recent progress in quantum error correction and hardware stability are signs that the field is moving toward practical systems. The next milestone, he said, will be the development of error-corrected quantum machines capable of addressing scientifically relevant problems, likely within the next few years.
For the HPC community, the Genesis Mission represents a continuation of a familiar model where large-scale computing infrastructure is directed toward solving some of the country’s most critical scientific and engineering challenges. Gil closed the discussion with a call for participation, framing Genesis as a national effort rather than a DOE-only program. The expectation is that its success will depend on how effectively the existing ecosystem, including labs, universities and industry, can coordinate around shared goals. If the model works, it could redefine how large-scale science is conducted in the U.S., with AI serving as a core part of the scientific method.
“This is the flagship national science and engineering initiative to transform the practice of our fields with the computing revolution and with AI, building a platform to accelerate discovery,” Gil said. “This is just the beginning. But we have to launch these teams, engage and make the Genesis mission your own. This is not a DOE story alone. This is a story for the whole country, an Apollo moment.”
Related
AI for Science, AI infrastructure, Dario Gil, department of energy, DOE, genesis mission, GTC 2026, high performance computing, HPC, Ian Buck, national laboratories, NVIDIA, quantum computing, scientific computing, Supercomputing
ORNL Introduces ‘Photon’ Framework for Accelerating AI Vulnerability Discovery on Frontier
March 25, 2026 — Oak Ridge National Laboratory’s Center for Artificial Intelligence Security Research (CAISER) is…
Gartner Forecasts 90% Drop in LLM Inference Costs by 2030
STAMFORD, Conn., March 25, 2026 — By 2030, performing inference on a large language model (LLM)…
Dario Gil at GTC: DOE’s Genesis Mission Moving from Vision to Execution
At Nvidia’s GTC conference last week, a fireside chat between Nvidia’s Ian Buck and U.S. Department of Energy Under…
Seven Reasons IBM’s Storage Chief Is Bullish on AI
As we sat down for an interview in the GTC Expo last week, it was…
Autoscience Secures New Funding to Scale Its Autonomous AI Research Lab
Autoscience, an applied research lab based in San Mateo, CA, has raised $14M in seed…
Alzheimer’s Disease Data Initiative Doubles $1M Prize Competition for Agentic AI Solutions
COPENHAGEN, Denmark, March 25, 2026 — The Alzheimer’s Disease Data Initiative (AD Data Initiative) has…
