This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| International Exascale Software Project | |
|---|---|
| Name | International Exascale Software Project |
| Abbreviation | IESP |
| Formation | 2009 |
| Type | International consortium |
| Purpose | Software infrastructure for exascale computing |
| Headquarters | International |
| Region served | Worldwide |
| Website | none |
International Exascale Software Project The International Exascale Software Project is an international consortium formed to coordinate software research and development for exascale supercomputing. The project engages with national laboratories such as Argonne National Laboratory, Lawrence Livermore National Laboratory, Oak Ridge National Laboratory, with universities including Massachusetts Institute of Technology, Stanford University, University of Cambridge, and with industry partners such as Intel, IBM, NVIDIA, AMD, and Cray Inc.. Its remit interacts with initiatives like Exascale Computing Project, European High Performance Computing Joint Undertaking, PRACE, and TOP500 efforts.
The project synthesizes expertise from entities like National Aeronautics and Space Administration, Department of Energy, European Commission, Riken, CERN, to address software ecosystems for systems exemplified by Summit (supercomputer), Fugaku, Frontier (supercomputer), and experimental platforms like Aurora (supercomputer). It considers runtime systems employed by OpenMP, MPI, CUDA, and OpenCL stacks while interfacing with standards bodies such as IEEE, ISO, The Open Group, and Khronos Group.
Origins link to collaborations among United States Department of Energy, European Union, and Japanese research agencies including METI and MEXT, and to projects like ASCR programs and Sequoia (supercomputer). Early workshops convened contributors from Sandia National Laboratories, Los Alamos National Laboratory, National Institute of Standards and Technology, Barcelona Supercomputing Center, and Rutherford Appleton Laboratory. Milestones reference coordination with DOE INCITE, HPC Advisory Council, and conferences such as Supercomputing Conference, International Conference for High Performance Computing, Networking, Storage and Analysis, ISC High Performance. The development timeline paralleled hardware advances at Intel Xeon Phi, NVIDIA Tesla, ARM server initiatives, and procurement programs by European Investment Bank and national ministries.
Primary objectives include enabling scalable scientific applications for codes like LAMMPS, GROMACS, VASP, Quantum ESPRESSO, and OpenFOAM on exascale platforms. It aims to coordinate middleware advances in HDF5, NetCDF, ADIOS, and in situ visualization tools such as ParaView and VisIt while aligning with verification programs linked to DOE ASC Program, European Centre for Medium-Range Weather Forecasts, and climate modeling efforts by Hadley Centre. The project sets targets for parallel I/O, fault tolerance using concepts developed in Checkpoint/Restart, energy-aware scheduling influenced by Green500 metrics, and performance portability promoted by Kokkos and RAJA.
Governance drew representatives from institutions like National Supercomputing Centre entities including NCSA, academic centers such as University of Illinois Urbana-Champaign, ETH Zurich, and corporate research labs like Microsoft Research, Google Research, Facebook AI Research, HPE, and Oracle. Working groups paralleled bodies such as OpenMP ARB, MPI Forum, ACM SIGARCH, IEEE Computer Society, and collaborated with consortia like OpenHPC and RDA. Advisory panels incorporated leaders from Cray Research heritage, contributors from Fujitsu, Toshiba, NEC Corporation, and user communities represented by projects like Human Genome Project and Square Kilometre Array.
The software stack emphasized toolchains involving GCC, LLVM, Intel C++ Compiler, and debuggers such as GDB and TotalView (debugger), performance tools including TAU Performance System, HPCToolkit, VTune Amplifier, and tracing frameworks like LTTng. Libraries targeted include BLAS, LAPACK, ScaLAPACK, PETSc, Trilinos, and solvers like Hypre and MLPACK. The project integrated resource managers and schedulers such as SLURM, PBS Professional, LSF, and incorporated containerization trends via Docker, Singularity (software), and orchestration references to Kubernetes. Security and provenance work drew from OAuth, OpenID, and data management principles practiced by The Research Data Alliance.
The initiative partnered with national programs including US Exascale Computing Project, EU EuroHPC, Japan's RIKEN Center for Computational Science, and research infrastructures like PRACE, GENCI, CSCS (Swiss National Supercomputing Centre), and HPC Wales. It engaged vendors such as NVIDIA Corporation, Intel Corporation, Advanced Micro Devices, Hewlett Packard Enterprise, and system integrators like Atos. Academic collaborations spanned University of California, Berkeley, University of Texas at Austin, Purdue University, Columbia University (New York), and domain science partnerships included NOAA, NASA Ames Research Center, European Space Agency, National Oceanography Centre, and projects like IPCC assessments.
Outcomes influenced standards adoption at MPI Forum, improvements in application readiness for platforms exemplified by Titan (supercomputer), and informed procurement decisions by agencies like DOE Office of Science, European Commission Horizon 2020, and national research councils. The project’s work seeded contributions to community codes such as OpenMP, MPI, HDF5, and shaped curricula at institutions like Carnegie Mellon University, Imperial College London, and Tsinghua University. Its legacy persists in successor initiatives including Exascale Computing Project, EuroHPC JU, and in the ecosystem surrounding supercomputers like Frontier (supercomputer), Fugaku, and next-generation architectures from NVIDIA and AMD.
Category:Supercomputing