The Office of Science of the U.S. Department of Energy, through its program Innovative and Novel Computational Impact on Theory and Experiment, or INCITE, has announced supercomputing access allocations to 51 high-impact computational science projects by 2022. These awards, which represent transformative progress in pursuing science and engineering account for 60 percent of the time available on the leadership-class supercomputers at DOE’s Argonne and Oak Ridge National Laboratories.
The projects will support a wide range of large-scale research campaigns to advance knowledge in areas ranging from astrophysics to sustainable energy technologies to materials design and discovery.
Jointly administered by the Argonne Leadership Computing Facility and the Oak Ridge Leadership Computing Facility, the INCITE program is the primary means by which the facilities fulfill their mission to advance open science by providing the scientific community with access to their powerful supercomputing resources. The ALCF and OLCF are user facilities of the DOE Office of Science.
The ALCF systems include Theta, a 16-petaflop Cray XC40 system backed by NVIDIA A100 GPUs, and Polaris, a 44-petaflop HPE Apollo 6500 Gen10+ testbed system that will provide scientists and application developers with a platform to code for Aurora, Argonne’s upcoming Intel -HPE exascale supercomputer. The OLCF’s flagship system is the 200 petaflop Summit, an IBM AC922 machine that debuted in 2018 as the most powerful computer in the world. Summit is also a testing ground for researchers as they prepare to run their codes on Frontier, the OLCF’s newest supercomputer, a 1.5+ exaflop HPE Cray EX system currently being installed. Time on Frontier will be allocated in the next INCITE cycle.
“INCITE assignments really serve as a gauge for the next frontiers in advanced computing. This year’s class features a diverse portfolio of ambitious research campaigns representing the most advanced techniques in high-performance computing to support a wide range of both applied and basic research,” said Gina Tourassi, director of the National Center for Computational Sciences, which houses the OLCF. “We are proud to provide full access to the world’s most powerful systems to our users who are at the forefront of their scientific fields.”
Open to any researcher or research organization in the world with a computationally intensive project, INCITE’s application process is highly competitive. Over a four-month period, INCITE proposals are assessed by 11 peer-review panels composed of international experts, with each panel representing a different scientific discipline. The proposals are also assessed at a technical level by each computing facility for compute readiness and the scalability of the project code and its algorithms. Based on these recommendations, the INCITE Prize Committee makes its final selection. This year, the committee received a total of 121 proposals, with researchers requesting more than 120,000,000 node hours for all three systems.
“Computers that can support both very powerful simulations and large-scale AI computations will provide researchers with a wider range of solutions to their problems,” said Michael E. Papka, director of Argonne Leadership Computing Facility. “Polaris is one such architecture that helps prepare critical workloads for future exascale systems. Today’s INCITE workloads on Polaris will become tomorrow’s Aurora workloads.”
The INCITE Committee added a new position in 2022 by devoting 10% of the allocable time to early careers. This track was launched to encourage the next generation of high-performance computing researchers. Researchers were able to apply within 10 years of earning their PhD, and nine early career projects were honored.
“We were very excited about the response to INCITE’s new early career path,” said Katherine Riley, INCITE program manager and ALCF director of science. “We look forward to working with a new generation of DOE supercomputer users to help them achieve breakthroughs in a number of key research areas ranging from drug discovery to climate modeling to hypersonic flight.”
Highlights of the 2021 allocations include:
Luca Comisso, an early career researcher from Columbia University, received 1,000,000 Theta node hours to develop first-principle simulations of black hole accretion currents and coronae.
Raquel Dias, an early career researcher at the Scripps Research Institute, received 690,000 Summit node hours to explore deep learning for accurate and cost-effective imputation of genotypes at the whole genome level.
ORNL’s Joshua New received 800,000 Theta node hours for modeling energy use in the country’s buildings.
Anastassia Alexandrova of the University of California Los Angeles received 2,000,000 Theta node hours to study heterogeneous catalysis as a collective phenomenon within dynamic ensembles of states.
Eric Nielsen of NASA’s Langley Research Center received 690,000 node hours from the summit to perform a multiphysical approach for guided simulations of the human-scale descent of Mars.
General Atomics’ Emily Belli received 540,000 Summit node hours to better understand turbulent transport in DT fusion plasmas.