Los Alamos National Laboratory

Los Alamos National Laboratory (LANL) is a US Department of Energy National Nuclear Security Administration (NNSA) national laboratory located in Los Alamos, New Mexico, founded in 1943 as Project Y of the Manhattan Project under the scientific leadership of J. Robert Oppenheimer. The laboratory operates the Venado NVIDIA Grace Hopper supercomputer (delivered 2024), the Crossroads NNSA Advanced Simulation and Computing system, and a portfolio of AI-for-science and AI-for-security research programs spanning nuclear stockpile stewardship, materials science, biology and genomics, climate modeling, and applied AI for national-security mission-critical applications. LANL is operated by Triad National Security LLC under contract to the National Nuclear Security Administration. As of April 2026, LANL is one of the principal US Department of Energy national laboratories by AI-for-science research output, with the longest continuous laboratory operating history of the NNSA-laboratory triad.

At a glance

• Founded: 1943 as Project Y of the Manhattan Project under J. Robert Oppenheimer.
• Status: US Department of Energy National Nuclear Security Administration (NNSA) national laboratory. Operated by Triad National Security LLC (a partnership of the University of California, Texas A&M University System, and Battelle Memorial Institute).
• Funding: US federal funding through the National Nuclear Security Administration and the US Department of Energy. Approximately $4.7 billion in fiscal year 2024 operating budget.
• Director: Thom Mason, Director of Los Alamos National Laboratory since 2018.
• Other notable leadership: Senior research leadership across the Venado compute, NNSA weapons-program, AI for science, and AI for security organizations.
• Open weights: Yes, partial. Selected research outputs released open-source through GitHub and academic-publication channels; classified weapons-program research is not publicly released.
• Flagship outputs: Venado NVIDIA Grace Hopper supercomputer; published AI-for-science research output across materials, biology, climate, and stockpile-stewardship simulation.

Origins

LANL was founded in 1943 at a remote mesa-top site in northern New Mexico as Project Y of the Manhattan Project, under the scientific leadership of J. Robert Oppenheimer and the military leadership of General Leslie Groves. The founding mission was the design and engineering of the first atomic weapons; the laboratory produced the Trinity test in July 1945 and the Hiroshima and Nagasaki weapons in August 1945.

The 1945 to 1985 period built LANL's principal research programs: nuclear weapons design (the laboratory's principal mission across the Cold War), high-performance computing (the early Cray and CDC platforms), inertial-confinement fusion research (predating the National Ignition Facility at Lawrence Livermore), space-and-planetary science (the laboratory's contributions to spacecraft RTG power systems), and applied physics across plasma, materials, and biology research. The lab's contributions to the early supercomputing era shaped the modern US national-laboratory compute infrastructure.

The 1990 to 2020 period saw LANL transition through the post-Cold War stockpile-stewardship era, with the NNSA Advanced Simulation and Computing program (ASCI from 1996 to 2002, ASC subsequently) anchoring the laboratory's compute-infrastructure investment and the simulation capabilities required to maintain the US nuclear stockpile without nuclear-explosive testing. The Trinity, Crossroads, and Roadrunner supercomputer generations occupied successive top positions in NNSA classified compute capability.

The 2020 to 2026 period saw LANL transition into AI-for-science and AI-for-security research at scale. The Venado NVIDIA Grace Hopper supercomputer delivered in 2024 brought NVIDIA's combined CPU-GPU APU architecture to the laboratory's research workloads, with applications across materials simulation, biology and genomics, climate modeling, and AI-for-security mission-critical applications. The laboratory's published AI-for-science research output through 2024 and 2025 spanned generative chemistry for materials discovery, AI-augmented stockpile-stewardship simulation, and AI-for-security applications including drone swarming, signal processing, and computer-network defense.

Mission and strategy

LANL's stated mission is to solve national-security challenges through scientific excellence. The strategy combines three principal threads. First, the classified NNSA stockpile-stewardship and weapons-program research that anchors the laboratory's principal mission. Second, open AI-for-science research across materials, biology, climate, and high-performance simulation. Third, AI-for-security research applied to mission-critical national-security applications including signal processing, computer-network defense, and autonomous-systems applications.

The strategic premise of the AI-for-science and AI-for-security programs is that frontier-scale AI capability applied to scientific simulation and to mission-critical national-security applications produces capability gains that academic and commercial AI labs cannot fully replicate within the classified-information environment.

Models and products

• Venado supercomputer. NVIDIA Grace Hopper APU-based supercomputer delivered 2024. Targeted at materials simulation, biology and genomics, climate modeling, and AI-for-security workloads.
• Crossroads. NNSA Advanced Simulation and Computing (ASC) system used for classified stockpile-stewardship simulation.
• AI-for-science research outputs. Published research across materials discovery, biology and genomics, climate modeling, and high-performance simulation.
• AI-for-security research outputs. Applied AI research for mission-critical national-security applications.

The distribution channels span US federal government NNSA classified workloads, US Department of Energy Office of Science unclassified research allocations, academic-research collaborations, and selected industrial-partnership engagements.

Benchmarks and standing

LANL's evaluation framework focuses on classified weapons-program performance, AI-for-science publication output, and supercomputer benchmarks. The laboratory's published research output through 2024 and 2025 has been characterized in industry coverage as one of the principal US national-laboratory AI-for-science research bodies.

Leadership

As of April 2026, LANL's senior leadership includes:

• Thom Mason, Director of Los Alamos National Laboratory since 2018.
• Senior research leadership across the Venado compute, NNSA weapons-program, AI for science, and AI for security organizations.

The Triad National Security LLC management partnership includes the University of California, Texas A&M University System, and Battelle Memorial Institute.

Funding and backers

US federal funding through the National Nuclear Security Administration and the US Department of Energy Office of Science. Approximately $4.7 billion in fiscal year 2024 operating budget. Operated as a federally funded research and development center (FFRDC) under the management of Triad National Security LLC.

Industry position

LANL occupies a distinctive position as the longest continuously operating US national laboratory and one of the principal NNSA laboratories by AI-for-science and AI-for-security research output. The combination of the Manhattan Project founding heritage, the Venado supercomputer, the long-running classified stockpile-stewardship base, and the published AI-for-science research output produces a profile that no peer US national laboratory matches at the same combination of attributes. Industry coverage has consistently characterized LANL alongside Lawrence Livermore and Sandia as the NNSA-laboratory triad.

Competitive landscape

• Lawrence Livermore National Laboratory, Sandia National Laboratories. NNSA-laboratory peers with overlapping classified weapons-program responsibility.
• Oak Ridge National Laboratory, Argonne National Laboratory, Lawrence Berkeley National Laboratory. US Department of Energy Office of Science peers with overlapping AI-for-science research focus.
• NSF AI Institutes, DARPA AI Next, DOE INCITE. US federal AI research-funding peer programs.
• NASA Frontier Development Lab. US federal AI-for-science research peer with a different (sprint-based) operational model.

Outlook

• Continued Venado utilization across AI-for-science and AI-for-security workloads through 2026 to 2027.
• Continued NNSA stockpile-stewardship simulation capability investment.
• Continued AI-for-science research output across materials, biology, climate, and stockpile-stewardship simulation.
• Continued AI-for-security applied research for mission-critical national-security applications.

Sources

• Los Alamos National Laboratory official site. Lab reference.
• Venado supercomputer announcement. Compute infrastructure reference.
• J. Robert Oppenheimer Wikipedia. Founding scientific director reference.
• NNSA national laboratories overview. Mission reference.