Nearly 100 scientists to share Argonne’s work in exascale, computing software, artificial intelligence methods and more.
Researchers from the U.S. Department of Energy’s (DOE) Argonne National Laboratory will highlight their most recent accomplishments in simulations, algorithm development, artificial intelligence (AI) and software at SC23, the International Conference for High Performance Computing, Networking, Storage and Analysis, taking place in Denver Nov. 12-16.
Rachana Ananthakrishnan and Tom Uram will talk about Nexus, a pioneering approach to connecting experimental, computing, and storage facilities at Argonne and beyond into an interconnected scientific infrastructure, in a featured talk at the DOE’s exhibit booth. Argonne researchers will also spotlight the Aurora exascale supercomputer at the Intel exhibit booth, including Aurora’s impact on the developer community, exascale application readiness and a demo of Aurora early science work using in-situ visualization.
Some notable Argonne activities are highlighted below. For the full schedule of conference participation, visit Argonne’s SC23 website.
Valerie Taylor, a Distinguished Fellow and director of Argonne’s Mathematics and Computer Science division, will moderate the SC23 opening plenary “I Am HPC: Impact and Future Directions.” Taylor’s research focuses on performance analysis and modeling of parallel scientific applications, and she is also CEO and president of the Center for Minorities and People with Disabilities in Information Technology. She will be joined by thought leaders for a lively discussion of the social impact of high performance computing (HPC) in the context of leading scientific and technical achievements and resources.
Argonne researchers are part of two teams that are finalists for this prestigious award. A team from Argonne, DOE’s Oak Ridge National Laboratory, Penn State University and the University of Illinois at Urbana-Champaign is a Gordon Bell Prize finalist for “Exascale Multiphysics Nuclear Reactor Simulations for Advanced Designs,” the first simulation of a fully coupled, fully resolved nuclear reactor core. Key to the innovative simulation by the research team is the coupling of NekRS, a spectral element code developed at Argonne for computational fluid dynamics simulations, and Shift, a high-fidelity Monte Carlo neutron transport code.
Argonne researchers are also part of a second multi-lab team led by DOE’s Sandia National Laboratories that is nominated for the new Gordon Bell Prize for Climate Modeling. The Energy Exascale Earth System Model (E3SM) team ran an unprecedented high-resolution global atmosphere model on the Frontier exascale supercomputer. Called the Simple Cloud Resolving E3SM Atmosphere Model, it provides a more accurate representation of cloud processes and their impact on climate change and weather patterns.
Superconducting digital computing (SDC) offers the benefits of reducing latency and energy disparity and has significant potential to preserve performance scaling for a wide range of high-performance computing applications. Rick Stevens, associate laboratory director of Computing, Environment and Life Sciences at Argonne, will participate in a panel focusing on how SDC may be used in practice in future systems and the research and development needed to realize that potential.
Murali Emani of the Argonne Leadership Computing Facility (ALCF), a DOE Office of Science user facility, will join colleagues to lead the fourth annual workshop on the use of AI/machine learning (ML) in scientific applications. Computer scientists and domain scientists from government, academia and industry will share advances, introduce new scientific application problems, and stimulate tools and infrastructures.
Argonne researchers will present several sessions at this workshop chaired by the ALCF’s Scott Parker. Esteban Rangel, Adrian Pope and Nicholas Frontiere will present work that evaluates the performance portability of a SYCL implementation of a large-scale cosmology application running on GPUs from three different vendors. And Geng Liu, Saumil Patel, Silvio Rizzi, Victor Mateevitsi and Joseph Insley will present a performance evaluation of heterogeneous GPU programming frameworks for hemodynamic simulations.
Argonne’s David Martin will co-host a panel discussion with industry representatives on how DOE’s Exascale Computing Project and the move to exascale computing is impacting industry’s current and planned use of high performance computing in saving energy, boosting competitiveness and building global technology leadership.
Now in its fifth year, the Extreme-Scale Experiment-in-the-Loop Computing (XLOOP) workshop focuses on the intersection of HPC and large-scale experimental science. Chaired by Argonne’s Justin Wozniak and Nicholas Schwarz, the XLOOP workshop will feature papers and presentations detailing how advances in computing, including new AI data transfer methods, are helping to accelerate discoveries by integrating HPC resources with observational and experimental facilities.
Argonne’s research in quantum information science — from theory and algorithms to simulations and modeling of quantum systems — will be featured when Yuri Alexeev co-chairs the Fourth International Workshop on Quantum Computing Software. He will also present a paper with Ruslan Shaydulin and others on “fast simulation of high-depth QAOA circuits” and join William Berquist and others in a talk on a scalable quantum architecture search package.
Argonne’s Lois Curfman McInnes will give an invited presentation on “Broadening Participation in HPC: Together We Can Change the World.” She will discuss how achievements in HPC drive advances in science and technology and how collaborative efforts, including those within ECP, are vital for creating a diverse and inclusive HPC workforce that can tackle big problems and change the world.