Two research teams from the U.S. Department of Energy’s (DOE) Argonne National Laboratory received Best Paper awards at the International Conference for High Performance Computing, Networking, Storage, and Analysis (SC24), held in November in Atlanta.

The teams were recognized for their work to advance in situ visualization capabilities and to provide performance benchmarking data on the Intel GPUs that power Aurora, the exascale supercomputer housed at the Argonne Leadership Computing Facility (ALCF). The ALCF is a DOE Office of Science user facility. 

The visualization study, “Bridging Gaps in Simulation Analysis through a General Purpose, Bidirectional Steering Interface with Ascent,” won the Best Paper award at SC24’s In Situ Infrastructures for Enabling Extreme-Scale Analysis and Visualization (ISAV 2024) Workshop. The paper was authored by Argonne’s Victor Mateevitsi, Silvio Rizzi, Joseph Insley, Michael Papka, Thomas Marrinan, and Dimitrios Fytanidis; other authors were Andres Sewell and Steve Petruzza of Utah State University and Cyrus Harrison and Nicole Marsaglia of DOE’s Lawrence Livermore National Laboratory. Sewell, the lead author, has been a summer student at the ALCF for the past two years.

Their paper introduces a framework for adding interactive, human-in-the-loop steering controls to existing simulation codes. This capability allows scientists to pause, adjust, and resume large-scale simulations without starting over. 

“This new interface gives researchers more flexibility and control over their simulations, making it possible to adapt them on the fly in ways that weren’t practical before,” said Mateevitsi, an ALCF assistant computer scientist. “We’re honored to receive the Best Paper award for this study, which was driven by our summer student, Andres.”

For the Intel GPU evaluation, an Argonne-led team received the Best Paper award for “Ponte Vecchio Across the Atlantic: Single-Node Benchmarking of Two Intel GPU Systems” at SC24’s Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems (PMBS24) Workshop. The paper was authored by Argonne’s Thomas Applencourt, Servesh Muralidharan, Colleen Bertoni, JaeHyuk Kwack, Ye Luo, Esteban Rangel, John Tramm, and Yasaman Ghadar; other authors were Aditya Sadawarte and Tom Deakin of the University of Bristol and Arjen Tamerus and Christopher Edsall of the University of Cambridge.

Their work provides micro-benchmarking data from applications running on two supercomputers powered by the Intel GPUs: ALCF's Aurora exascale system and University of Cambridge's Dawn system. 

“Our goal was to characterize single-node performance on the systems, providing detailed performance benchmark results to help guide developers working to optimize applications to run on the new Intel GPUs,” said Applencourt, an ALCF computational scientist.