National Supercomputing Centre

National Supercomputing Centre
Name	National Supercomputing Centre
Established	2000s
Location	Various

National Supercomputing Centre

The National Supercomputing Centre is a high-performance computing institution that provides large-scale computational resources for scientific, industrial, and governmental use. It houses petascale and exascale-class systems that support research in fields such as climate science, bioinformatics, aerospace engineering, and materials science. The centre operates partnerships with universities, national laboratories, ministries, and international consortia to advance computational infrastructure and algorithm development.

Introduction

The centre offers tiers of compute, storage, and visualization resources comparable to installations like Oak Ridge National Laboratory, Lawrence Livermore National Laboratory, Argonne National Laboratory, Los Alamos National Laboratory, and CERN. It hosts systems from vendors such as IBM, Cray, Hewlett Packard Enterprise, NVIDIA, and Intel and interoperates with networks like ESnet, GEANT, Internet2, and GLORIAD. The centre supports codes and frameworks including MPI (Message Passing Interface), OpenMP, CUDA, OpenACC, and HDF5, and collaborates with projects such as Apache Hadoop, TensorFlow, PyTorch, Jupyter Notebook, and Singularity (container).

History and Development

Origins of the centre trace to national initiatives inspired by programs at National Science Foundation, European Commission, Japan Science and Technology Agency, Deutsches Forschungszentrum für Künstliche Intelligenz, and investments following models from Franco-German cooperation and Singapore National Research Foundation. Early procurement cycles referenced machines like Blue Gene and Cray XT and partnerships with laboratories similar to Princeton Plasma Physics Laboratory and Fermi National Accelerator Laboratory. Upgrades followed the trajectories of milestones exemplified by Top500 listings, the Green500 energy-efficiency chart, and standards from OpenStack and POSIX communities. Strategic plans echoed roadmaps from Horizon 2020, U.S. Department of Energy, Riken, and Compute Canada.

Facilities and Architecture

The facility comprises modular data halls, high-density racks, liquid-cooling plants, and uninterruptible power systems modeled on deployments at Jülich Research Centre, EPCC, Barcelona Supercomputing Center, and National Centre for Supercomputing Applications. Networking uses fabrics like InfiniBand, Omni-Path, Ethernet (network) switches from Cisco Systems, Arista Networks, and accelerators from AMD and ARM Holdings. Storage tiers implement technologies like parallel file systems such as Lustre, GPFS, and object stores inspired by Ceph. Security architecture references best practices from NIST, ISO/IEC 27001, and compliance frameworks similar to GDPR in handling research data.

Research and Applications

Researchers run simulations and analyses in domains related to Intergovernmental Panel on Climate Change, Human Genome Project, CRISPR, NASA, European Space Agency, Boeing, Airbus, and Siemens. Projects include climate modeling comparable to Hadley Centre ensembles, cosmology simulations akin to Millennium Run, fusion modeling like ITER support work, and drug-discovery pipelines parallel to efforts at Pfizer, Roche, and GlaxoSmithKline. Machine learning workloads integrate models and benchmarks such as ImageNet, BERT, GPT (language model), and scientific software like LAMMPS, GROMACS, VASP, ANSYS, and OpenFOAM.

Governance and Funding

Governance structures mirror frameworks seen at European Research Council, National Institutes of Health, Department of Energy (United States), Ministry of Science and Technology (country), and Wellcome Trust. Funding sources comprise national research budgets, competitive grants from entities like Horizon Europe, National Science Foundation, Engineering and Physical Sciences Research Council, philanthropic foundations such as Gordon and Betty Moore Foundation, and industry consortiums including OpenAI, Microsoft Research, and Google Research. Advisory boards include representatives from University of Cambridge, Massachusetts Institute of Technology, Stanford University, University of Tokyo, and national academies like Royal Society and National Academy of Sciences.

Collaborations and Partnerships

The centre maintains academic partnerships with institutions such as Imperial College London, University of Oxford, ETH Zurich, Tsinghua University, Peking University, University of Toronto, University of Melbourne, KTH Royal Institute of Technology, and Seoul National University. It participates in international consortia including PRACE, XSEDE, Gaia (spacecraft) data pipelines, SKA preparatory efforts, and joint ventures with industry partners like Intel Corporation, NVIDIA Corporation, AMD (company), ARM Limited, Dell Technologies, and Lenovo. Memoranda of understanding reference collaborative models practiced by European Grid Infrastructure and Open Science Grid.

Outreach and Education

Educational programs include training workshops, summer schools, and fellowships modeled after initiatives from SDSC, NERSC, Leibniz Supercomputing Centre, and Compute Ontario. Outreach engages students and educators through hackathons, internships, and curriculum resources linked with Coursera, edX, Khan Academy, and university extension programs from Columbia University, Princeton University, and University of California, Berkeley. Public engagement leverages exhibitions and briefings similar to those held at Science Museum, London, Smithsonian Institution, and science festivals like World Science Festival.

Category:Supercomputer sites