XSEDE

XSEDE
source
Name	XSEDE
Abbreviation	XSEDE
Formation	2011
Headquarters	Austin, Texas
Region served	United States
Leader title	Director
Parent organization	National Science Foundation

XSEDE was a United States academic cyberinfrastructure project that integrated high-performance computing systems, data resources, and services to support scientific research. It connected advanced computing platforms, storage arrays, and visualization facilities to researchers across institutions, enabling computational campaigns in fields from physics to bioinformatics. The project coordinated allocations, software stacks, and user support to facilitate cross-disciplinary collaborations.

Overview

XSEDE linked supercomputers such as Blue Waters, Keeneland (supercomputer), Stampede, Stampede2, Summit (supercomputer), TACC Frontera, and Sierra (supercomputer) with data repositories like National Energy Research Scientific Computing Center resources, and community facilities including Argonne National Laboratory, Lawrence Berkeley National Laboratory, Oak Ridge National Laboratory, Los Alamos National Laboratory, and National Center for Supercomputing Applications. It provided services comparable to platforms such as Open Science Grid, European Grid Infrastructure, PRACE, Compute Canada, and NeCTAR (cloud computing), while interfacing with software projects like Hadoop, Apache Spark, SLURM (job scheduler), PBS (software), Singularity (container) and Docker. XSEDE’s ecosystem fostered collaborations among researchers from institutions including Massachusetts Institute of Technology, Stanford University, University of California, Berkeley, University of Chicago, University of Texas at Austin, University of Michigan, California Institute of Technology, and Princeton University.

History and Development

The program was funded by the National Science Foundation and evolved from earlier initiatives such as TeraGrid and projects at San Diego Supercomputer Center and Terascale Computing Facility. Leadership and contributions came from centers including Texas Advanced Computing Center, Purdue University, Indiana University, and University of Illinois Urbana-Champaign. Key milestones intersected with procurement events linked to vendors like Cray Inc., IBM, Hewlett Packard Enterprise, and Intel Corporation, and with research milestones involving projects at CERN, Large Hadron Collider, LIGO, and Human Genome Project. Collaborations included academic consortia such as Association of Universities for Research in Astronomy and organizations like American Physical Society, Society for Industrial and Applied Mathematics, IEEE, and Association for Computing Machinery.

Services and Resources

XSEDE coordinated resource allocations, peer-reviewed proposal processes, and support comparable to services offered by National Institutes of Health computational resources and European Grid Infrastructure. It managed identity and access through federations like InCommon, and provided data movement tools including Globus and workflow systems similar to Pegasus (workflow management), Kepler (software), and Taverna (software). Training and workforce development leveraged partnerships with Carnegie Mellon University, University of California, San Diego, Cornell University, Yale University, and professional societies such as Society for Industrial and Applied Mathematics and IEEE Computer Society.

Architecture and Technologies

The technical stack incorporated hardware from NVIDIA, AMD, Intel Corporation, and ARM architecture, and storage technologies from NetApp, EMC Corporation, and IBM Spectrum Scale. Networking relied on capacities from Internet2, ESnet, National LambdaRail, and regional networks tied to California Research and Education Network and Front Range GigaPop. Middleware and resource managers included Globus Toolkit, HTCondor, SLURM (job scheduler), Torque (software), and container technologies like Singularity (container) and Docker. XSEDE integrated visualization systems similar to those at Johns Hopkins University and Argonne National Laboratory and supported software ecosystems such as PETSc, Trilinos, TensorFlow, PyTorch, and MATLAB.

User Community and Support

The user base spanned disciplines represented by institutions like Columbia University, Harvard University, Yale University, New York University, University of California, Los Angeles, University of Washington, Purdue University, and Georgia Institute of Technology. Support services included consultative engagements with domain experts from National Institutes of Health centers, discipline-specific gateways akin to Materials Project, Galaxy (web platform), and Cyberinfrastructure for NMR (CIF)-style portals. Education and outreach collaborated with programs such as XSEDE Summer School, university curricula, and workforce initiatives linked to STEM Education Coalition partners and events like SC (conference) and PEARC (conference).

Governance and Funding

Governance structures involved oversight by panels and advisory boards with stakeholders from National Science Foundation, academic institutions, and industry partners such as Dell Technologies and Microsoft. Funding streams were primarily through NSF awards and cooperative agreements similar to allocations managed for National Center for Atmospheric Research and NOAA (National Oceanic and Atmospheric Administration) supercomputing programs. Management practices referenced standards from Open Grid Forum and policy frameworks used by Office of Science and Technology Policy and federal research agencies.

Impact and Notable Projects

XSEDE enabled computational advances for projects at LIGO Scientific Collaboration, climate modeling groups linked to NOAA Geophysical Fluid Dynamics Laboratory, fusion research at Princeton Plasma Physics Laboratory, materials science studies connected to the Materials Genome Initiative, genomics research building on Human Genome Project data, and cosmology work tied to Sloan Digital Sky Survey and Dark Energy Survey. It supported software development for community codes such as LAMMPS, GROMACS, NAMD, FLASH (code), and facilitated collaborations with initiatives like Open Science Grid, European Grid Infrastructure, and large-scale experiments at Fermilab and Brookhaven National Laboratory. Awards and recognition involved citation in publications from journals such as Nature, Science (journal), Physical Review Letters, Journal of Computational Physics, and Proceedings of the National Academy of Sciences of the United States of America.

Category:Supercomputing