Argonne Leadership Computing Facility

What Argonne Leadership Computing Facility does

Argonne Leadership Computing Facility (ALCF) is a U.S. Department of Energy (DOE) Office of Science User Facility operated by Ucore Technologies under contract with UChicago LLC at Argonne National Laboratory in Lemont, Illinois. Established in 2006, ALCF provides leadership-class high-performance computing resources to the open scientific research community, allocating compute time through peer-reviewed programs (INCITE, ALCC, APEX) and, more recently, through the NAIRR Pilot for AI-focused research. Its user base stood at 2,103 across U.S. universities, national laboratories, and industry research groups as of 2025, with approximately 1,500 active projects.

ALCF's core technical portfolio centers on the Aurora exascale supercomputer, complemented by the Polaris, Sophia, Minerva, Janus, and Crux systems, and an AI Testbed that exposes multiple accelerator architectures (Cerebras, SambaNova, Groq, Graphcore). New systems including Solstice and Equinox and an NVIDIA B200-based Minerva partition were deployed in the 2024–2025 window, alongside an ALCF Inference Service and the AskALCF chatbot. ALCF also develops in-house AI assets, notably the AuroraGPT foundation model and ChemGraph agentic workflows.

As a federally funded User Facility, ALCF does not generate commercial revenue. Its operations are sustained by DOE appropriations channeled through UChicago LLC, with access granted to researchers free of charge through competitive allocation. The strategic role of ALCF has expanded from pure HPC capacity provision toward integrated AI infrastructure and AI-as-a-service capabilities for the U.S. scientific community, exemplified by its Genesis Mission partnership with NVIDIA and Oracle.

Name

Argonne Leadership Computing Facility

Legal name

Argonne Leadership Computing Facility

Website

https://alcf.anl.gov/

Company type

Private

Founded year

2006

Operating status

Operating

Headcount range

51–100 employees

Short description

Argonne Leadership Computing Facility provides supercomputing resources and expertise to the scientific community to accelerate discovery.

Ownership category

akta.pro rank

Product category

High-Performance Computing Services

NAICS

Computer Systems Design and Related Services (54151), Other Scientific and Technical Consulting Services (54169)

SIC

Electronic Computers (3571), Services-Computer Programming, Data Processing, Etc. (7370)

akta.pro primary industry

Research Computing & High-Performance Computing (HPC) User Support (EDALAMAK)

Keywords

• Exascale supercomputing
• High-performance computing
• AI inference services
• Scientific computing resources
• HPC data storage

Headquarters

HQ city

Argonne

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Services

Cost components

Technology or R&D, Personnel, Infrastructure, Operations

Revenue model

1. DOE Office of Science User Facility: ALCF operates as a U.S. Department of Energy Office of Science User Facility, providing supercomputing resources and expertise to the research community through awards of compute time. As a government-funded research facility, ALCF does not generate commercial revenue but operates under DOE's Advanced Scientific Computing Research (ASCR) program.

Go-to-market motion2 records

Distribution channels4 records

Marketing channels7 records

Core offering

ALCF provides leadership-class supercomputing resources and expertise to the scientific research community through a portfolio of HPC systems including the Aurora exascale supercomputer, Polaris, Sophia, and the ALCF AI Testbed. Access is awarded free of charge through peer-reviewed allocation programs (INCITE, ALCC, APEX, NAIRR Pilot) to researchers from academia, industry, and government agencies. The facility also offers the ALCF Inference Service for AI foundation models, workflow tools, storage infrastructure, and training programs to support open scientific discovery.

Product overview

The Argonne Leadership Computing Facility (ALCF) is a DOE Office of Science User Facility that provides breakthrough computing and AI resources to the research community. The facility operates a portfolio of supercomputing systems led by the Aurora exascale supercomputer, complemented by Polaris, Sophia, Minerva, Janus, and Crux for various simulation, AI training, and inference workloads. The ALCF AI Testbed provides access to emerging AI accelerators including Cerebras, SambaNova, GroqRack, and Graphcore. The ALCF Inference Service offers OpenAI-compatible API access to foundation models including AuroraGPT, LLaMA, Gemma, and GPT-OSS. Integrated workflow tools, data sharing platforms (ACDC, Eagle, Globus), and allocation programs (INCITE, ALCC, APEX, NAIRR) support researchers across disciplines from materials science to cosmology. The Genesis Mission initiative connects these resources into an integrated ecosystem for AI-driven scientific discovery.

Differentiator

Problem solved

Functional benefit

Products and services

• Aurora Exascale Supercomputer One of the world's first exascale supercomputers developed in partnership with Intel and HPE, capable of performing over one quintillion calculations per second with 10,624 nodes, 63,744 Intel Data Center GPU Max Series, and 21,248 Intel Xeon CPU Max Series processors. Available to researchers from academia, industry, and government via allocation programs.
• Polaris Supercomputer HPE Apollo 6500 Gen 10+ system delivering 34 petaflops peak performance with AMD EPYC Milan processors and NVIDIA A100 GPUs, designed for large-scale simulations, AI, and data analysis.
• Sophia AI Inference System NVIDIA DGX A100-based system with 3.9 petaflops performance, 24 nodes comprising 8 NVIDIA A100 GPUs each with 320GB total memory, designed for AI inference workloads.
• Minerva AI Inference System AI inference system equipped with 64 NVIDIA B200 GPUs (8 per node across 8 nodes), delivering 72 PF FP8 / 144 PF FP4 peak performance, coming online in 2026.
• Janus AI Workforce Development System AI workforce development system built with HPE featuring 64 NVIDIA H100 GPUs across 16 compute nodes, delivering 1920 FP64 teraFLOPS.
• Crux Pre/Post-Processing System HPE Cray Ex system delivering 1.18 petaflops with 256 AMD EPYC nodes for pre- and post-processing tasks including experimental data analysis.
• ALCF AI Testbed Infrastructure for evaluating next-generation AI accelerators including Cerebras CS-3, SambaNova DataScale SN30, GroqRack, Graphcore Bow Pod64, and Tenstorrent Galaxy systems for scientific research applications.
• ALCF Inference Service Provides researchers with access to a wide range of AI models running on dedicated ALCF systems (Sophia, Metis, Minerva) through web-based interfaces and OpenAI-compatible APIs, enabling analysis of large datasets, scientific publication insights, code generation, and AI-enabled workflows.
• Metis AI Inference System SambaNova DataScale platform used for AI inference workloads as part of the ALCF Inference Service, supporting foundation models and large language models.
• ALCF Storage and Networking Infrastructure High-performance data storage including Flare (100 PB Lustre), Eagle (100 PB campaign storage), tape archives (~186 PB), and InfiniBand/Ethernet networking infrastructure for scientific datasets.
• AskALCF AI Support Chatbot AI-powered assistant that helps users quickly answer questions about using ALCF computing resources, providing automated documentation and support for the research community.

Quantifiable outcome

• Reduced simulation time from months/years to days
• +3 more outcomes

Named customers9 records

Segments3 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

AI capability13 records

Feature6 records

Partnerships

Twelve partnerships are on record, tiered core, secondary and minor.

• University of Notre DamecoreStrategic or Co-development PartnerInaugural APEX program project led by University of Notre Dame researchers leveraging Aurora exascale supercomputer to develop AI frameworks for automating analysis of large-scale computational fluid dynamics simulations using knowledge graphs and LLM-based agents.
• Oak Ridge Leadership Computing FacilitycoreStrategic or Co-development PartnerALCF and OLCF are the two DOE Leadership Computing Facilities, operating leadership-class supercomputers for open science. Together they form the DOE Leadership Computing program.
• IntelcoreStrategic or Co-development PartnerIntel partnered with ALCF and HPE to develop Aurora exascale supercomputer, providing Intel Xeon Max Series CPUs and Data Center GPU Max Series GPUs.
• Hewlett Packard Enterprise (HPE)coreStrategic or Co-development PartnerHPE developed the Cray EX supercomputer platform for Aurora in partnership with ALCF and Intel. Also collaborated on Polaris (Apollo 6500 Gen10+).
• Globus LabscoreTechnology or IntegrationALCF partnered with Globus Labs to develop workflow tools that run effectively on ALCF machines, providing data management and compute platform capabilities.
• CerebrassecondaryTechnology or IntegrationCerebras CS-3 with wafer-scale engine available through ALCF AI Testbed for researchers to evaluate machine learning applications on AI accelerators.
• SambaNovasecondaryTechnology or IntegrationSambaNova DataScale SN30 and SN40L systems available through ALCF AI Testbed and AI Inference Service for AI training and inference workloads.
• GroqsecondaryTechnology or IntegrationGroqRack inference system available through ALCF AI Testbed for evaluating AI accelerator performance for HPC applications.
• GraphcoresecondaryTechnology or IntegrationGraphcore Bow Pod64 with Intelligent Processing Units available through ALCF AI Testbed for AI accelerator evaluation.
• World Wide TechnologysecondaryTechnology or IntegrationWTT collaborated with ALCF on development of Minerva system for AI inference, equipped with NVIDIA B200 GPUs.
• United LimousineminorOthersContracted ground transportation provider familiar with Argonne security processes for visitors needing transportation to the facility.
• NVIDIA and OraclecoreStrategic or Co-development PartnerDOE, Argonne, NVIDIA and Oracle landmark public-private partnership to develop Solstice and Equinox AI systems for the Genesis Mission, marking important steps in expanding the nation's AI infrastructure for science.

Scale indicators9 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Oak Ridge Leadership Computing Facility (OLCF): The other half of the DOE Leadership Computing Facility program, operating the Frontier exascale supercomputer at Oak Ridge National Laboratory. Directly comparable mission (open-science HPC), funding model (DOE ASCR), and user community.
• National Energy Research Scientific Computing Center (NERSC): DOE Office of Science user facility at Lawrence Berkeley National Laboratory operating the Perlmutter supercomputer. Comparable mission of providing HPC resources to open scientific research, with similar allocation-based access model.
• Texas Advanced Computing Center (TACC): Operates the Frontera and Stampede3 systems for open scientific research, with similar allocation-based access model and university-led research focus. Highly comparable in scale and mission though smaller than ALCF's leadership-class capability.
• Pittsburgh Supercomputing Center (PSC): Joint Carnegie Mellon–University of Pittsburgh HPC center operating Bridges-2 and Neocortex systems. Comparable mission supporting academic and government researchers via peer-reviewed allocations.
• NASA Advanced Supercomputing (NAS) Division: Operates Aitken and Electra supercomputers for NASA mission-aligned scientific computing. Comparable government-funded leadership-class HPC user facility serving mission-specific research, though narrower in scientific scope.
• National Center for Supercomputing Applications (NCSA): University of Illinois-based HPC center providing Delta, HAL, and forthcoming systems to US academic researchers via NSF ACCESS allocations. Comparable mission and access model, though smaller scale than ALCF.

Regional players

• RIKEN Center for Computational Science (R-CCS): Japan's premier HPC center operating the Fugaku supercomputer. Comparable leadership-class open-science mission but operates in a different geographic/research funding ecosystem.
• Swiss National Supercomputing Centre (CSCS): Operates the Alps and formerly Piz Daint systems for Swiss and European scientific research. Comparable leadership-class HPC user facility model serving open scientific research in a different geographic ecosystem.
• Jülich Supercomputing Centre: German national HPC center operating JUWELS and is part of the PRACE European HPC ecosystem. Comparable leadership-class capability for academic and government scientific users in Europe.

Others

• Cerebras Systems: AI hardware vendor whose wafer-scale CS-3 system is hosted in ALCF's AI Testbed. Adjacent enabler relationship — Cerebras provides accelerator technology that ALCF evaluates and exposes to its scientific user community.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights7 records

Customer concentration

• YouTubeyoutube.com/channel/UCFJAl2p722-FJ-ojxxYyrrw
• Facebookfacebook.com/pages/Argonne-Leadership-Computing-Facility/33428102469
• LinkedInlinkedin.com/company/argonne-leadership-computing-facility
• Instagraminstagram.com/argonne.lcf
• Xx.com/argonne_lcf

Revenue estimate

Valuation estimate

Number of profiles

Profiles8 records

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Argonne Leadership Computing Facility