History

The ALCF was established at Argonne National Laboratory in 2006 as part of a DOE initiative dedicated to enabling transformational advances in science and engineering by providing the most powerful supercomputers in the world for open scientific research.

Margaret K. Butler

Margaret Butler programmed the first digital computers at Argonne National Laboratory in the early 1950s, helped design subsequent ones, and contributed to simulations of nuclear power reactors. Image: Argonne National Laboratory

The facility has grown leaps and bounds since its inception, developing and deploying leading-edge computing systems with unprecedented capabilities, while fostering an ever-growing user community that is pushing the boundaries of fundamental discovery and understanding in a broad range of scientific disciplines.

Supported by the Advanced Scientific Computing Research (ASCR) program within DOE’s Office of Science, the ALCF is one half of the DOE Leadership Computing Facility, which deploys to diverse high-performance computer architectures for open science (one at Argonne and one at Oak Ridge National Laboratory).

ALCF Continues a Tradition of Computing Innovation

In 1949, Argonne physicists needed computers to solve enormously complex mathematical problems. But these kinds of computers were not yet available commercially, so they built their own. Thus began the Argonne tradition of innovation in providing leading computing resources to the scientific community.

The seedbed for such groundbreaking software as MPI, PETSc, PVFS, GridFTP, and Cobalt, and as an essential partner in the development of some of the world’s premier supercomputers, Argonne’s historic stronghold at the forefront of scientific computing is undisputed.

Notable Activities

2023

Supported 417 active projects, including 18 INCITE projects and 15 ALCC projects
Delivered 35.7 million node-hours of compute time
Supported 1,624 facility users
Sunspot became available to ESP and ECP reearchers
The ALCF AI Testbed gained Graphcore and GroqRack as production systems
Aurora was fully installed and entered the Top500 in the #2 spot with a partial system run
Hosted 34 training events

2022

Supported 385 active projects, including 21 INCITE projects and 16 ALCC projects
Delivered 35 million node-hours of compute time
Supported 1,538 facility users
Polaris and the ALCF AI Testbed’s Cerebras and SambaNova systems were deployed
ALCF resources and staff amongst the team to win the Gordon Bell Special Prize at SC22

2021

Supported 375 active projects, including 16 INCITE projects, 16 ALCC projects, and 4 ALCF Data Science Projects
Delivered 34 million node-hours of compute time
Supported 1,168 facility users
Polaris testbed supercomputer is delivered and installed at the ALCF

2020

Supported 369 active projects, including 14 INCITE projects, 36 ALCC projects, 6 ALCF Data Science Program projects, and 15 Aurora Early Science Program projects
Delivered 101.3 million node-hours of compute time (includes 3 months of Mira)
Supported 1,174 facility users
Through the CARES Act, Theta was augmented with 24 NVIDIA DGX A100 GPU nodes, increasing it to 15.6 petaflops

2019

Supported 413 active projects, including 29 INCITE projects, 32 ALCC projects, and 8 ALCF Data Science Program projects
Delivered 8.7 billion core-hours of science
Supported 1,246 facility users
Research on ALCF resources resulted in 275+ publications
ALCF Staff Members Contribute to SC19 Technology Challenge Award

2018

Supported 32 INCITE projects, 20 ALCC projects, and 8 ALCF Data Science Program projects
Expanded the Aurora Early Science Program to include data science and machine learning projects
Argonne hosted its first Quantum Computing Workshop
A team of Argonne/ALCF researchers received an R&D 100 award for the development of Darshan

2017

Theta entered production mode on July 1, 2017
Supported more than 350 active projects, including 30 INCITE projects and 24 ALCC projects
Launched Aurora Early Science Program with the selection of 10 simulation-based projects
Expanded the ALCF Data Science Program with the addition of four new projects
Research on ALCF resources results in 200+ publications

2016

Theta arrived
Kicked-off Argonne Data Science Program w/ 4 tier one projects
Supported 34 INCITE projects, 27 ALCC projects, and 160+ Director’s Discretionary projects
Research on ALCF resources resulted in 250+ publications

2015

Announced our next-generation supercomputers, Aurora and Theta
Kicked-off Theta Early Science Program w/ 6 tier one projects and 6 tier two projects
Launched Cooley, a new visualization and analysis cluster with nearly eight times the memory capacity of the facility’s previous system, Tukey
Supported 37 INCITE projects, 24 ALCC projects and 190+ Director’s Discretionary projects
Research on ALCF resources resulted in 160+ publications

2014

Supported 340+ active projects, including 40 INCITE projects and 18 ALCC projects
Supported 1,432 DOE facility users
Delivered 5.8 billion core-hours of compute time
Research on ALCF resources resulted in 150+ publications

2013

Supported more than 300 active projects, including 37 INCITE projects and 9 ALCC projects
Supported 1,150 DOE facility users
Mira, the ALCF’s Blue Gene/Q supercomputer, went into production on April 9, 2013
Intrepid, the ALCF's Blue Gene/P supercomputer, was decommissioned on December 31, 2013
Delivered 4.79 billion core-hours of compute time
Research on ALCF resources resulted in more than 150 publications
Two ALCF publications receive Best Paper Awards
Four ALCF staff papers accepted to SC13, including a finalist for the Gordon Bell Prize

2012

Supported 31 INCITE and 9 ALCC projects sponsored by the U.S. Department of Energy
Mira, the ALCF’s new Blue Gene/Q supercomputer, was delivered, installed and accepted
Mira was ranked the fourth fastest supercomputer in the world, according to the November 2012 TOP 500 list, and the second fastest supercomputer, according to the November 2012 Graph 500 list
16 Early Science Projects were first to gain access to Mira
Delivered 1.2 billion core-hours of science on Intrepid
Four ALCF staff members were part of a Gordon Bell finalist paper at SC12

2011

Supported 30 INCITE projects and 7 ALCC projects sponsored by the Department of Energy
Planned facility enhancements for Mira, the next-generation Blue Gene/Q supercomputer
Supported 16 Early Science projects
Prepared for Tukey, our visualization system that will be twice as fast as its predecessor, Eureka
Delivered 1.2 billion core-hours of science
Four ALCF staff members were part of a Gordon Bell finalist paper that won a Honorable Mention at SC11

2010

Began support of 35 INCITE projects and 10 ALCC projects sponsored by the Department of Energy
Signed contract for Mira, the next-generation Blue Gene/Q supercomputer
Selected 16 Early Science projects to run on the 10-petaflops Mira
Delivered more than 2 billion core-hours of science
Intrepid won the top spot in the first set of results published on the new Graph 500 benchmark at SC10

2009

Intrepid, a 557-teraflops Blue Gene/P system, went into production
Began support of 28 INCITE projects
Approved for 10 petaflops system to be delivered in 2012 timeframe
Began joint Argonne/NERSC Magellan cloud project
Delivered 897 million core-hours of science

2008

Increased to 20 INCITE projects
Begin support of Early Science and INCITE projects on Blue Gene/P
Intrepid (Blue Gene/P) won first place in RandomAccess and FFT benchmarks in HPC Challenge at SC08
Intrepid was initially ranked #4 on the Top500 list in June at ISC08

2007

Increased to 9 INCITE projects
Continued development projects
Held Next Generation Blue Gene workshop (June)
Installed 100-teraflops Blue Gene/P (Oct.-Nov.)
Accepted 100-teraflops Blue Gene/P (Dec.)

2006

The ALCF is officially established as a division at Argonne
Began production support of 6 INCITE projects
Continued code development and evaluation

2005

Installed 5-teraflops Blue Gene/L for evaluation

2004

Argonne and Oak Ridge are selected as the locations for the DOE Office of Science Leadership Computing Facilities after a competitive peer review of proposals
Formed the Blue Gene Consortium with IBM