The 2022 Gordon Bell Prize is awarded to plasma accelerator research

At the awards ceremony at SC22 in Dallas today, ACM presented the 2022 ACM Gordon Bell Prize to a team of researchers who used four major supercomputers – including the exascale Frontier system – to conduct extensive research into the plasma accelerator technologies.

The Gordon Bell Prize is awarded annually to SC22 to celebrate outstanding achievement in high-performance computing – specifically, according to the ACM, for “tracking progress over time in parallel computing, with particular emphasis on the award for innovation in the application of high performance”. performance computing to applications in science, engineering, and large-scale data analytics. Six finalists were selected earlier this year ahead of the selection of the winning team, which will receive a $10,000 prize courtesy of HPC luminary Gordon Bell.

On Tuesday, the six finalists presented their research to packed houses, with topics including atomic analysis of metallic materials; analytics for biomedical literature; plasma simulations for laser-based accelerators; geostatistics; earthquake modeling; and protein similarity searches.

Six major systems have fueled this research. OceanLight, the Chinese exascale system that powered last year’s Gordon Bell Prize-winning research, made another appearance – but this time it was met by research being conducted on America’s new exascale system, Oak Ridge National Lab’s Frontier, which debuted this summer. RIKEN’s Fugaku system and Oak Ridge’s Summit system also returned to the list (powering several finalist projects), along with NERSC’s Perlmutter system and KAUST’s Shaheen-2 system.

At the end of the day, however, only one of the teams can win the award – and when ACM President Cherri Pancake took the stage, the first thing she announced was that the 2022 ACM Gordon Bell Award was awarded to…

Pushing the boundaries in laser-based electron accelerator design with groundbreaking simulations of particles in mesh-refined cells on exascale-class supercomputers

Although the title of the article – which revolved around kinetic plasma simulations – winks at its use of Frontier, the team actually used four supercomputers: Frontier, Fugaku (RIKEN), Summit and Perlmutter. (NERSC), meaning this article used four of the top seven supercomputers on the most recent Top500 list. In an email to HPCwireJean-Luc Vay – senior scientist at Lawrence Berkeley National Lab – described the scientific leads of the research, which were conducted on Frontier (up to 8,192 knots), Fugaku (up to ~93,000 knots) and Summit ( up to 4,096 knots).

“Plasma accelerator technologies have the potential to deliver much more compact particle accelerators than existing ones, opening the door to exciting new applications in science, industry, security and safety. health,” Vay explained. “Harnessing the world’s most powerful supercomputers to drive research to make these complex machines a reality is so exciting for all of us.”

The abstract of the article is included below. Its authors include Luca Fedeli, Axel Huebl, France Boillod-Cerneux, Thomas Clark, Kevin Gott, Conrad Hillairet, Stephan Jaure, Adrien Leblanc, Remi Lehe, Andrew Myers, Christelle Piechurski, Mitsuhisa Sato, Neil Zaim, Weiqun Zhang, Jean-Luc Vay and Henry Vincent.

Summary: We present a massively parallel refined mesh (MR) PIC (Particle-In-Cell) code for optimized plasma kinetic simulations on Frontier, Fugaku, Summit, and Perlmutter supercomputers. Major innovations, implemented in the WarpX PIC code, include: (i) a three-tier parallelization strategy that has demonstrated performance portability and scalability across millions of A64FX cores and tens of thousands of AMD and Nvidia GPUs (ii) a breakthrough mesh refinement capability that provides between 1.5x to 4x savings in computing needs over the scientific case reported in this article, (iii) an efficient load balancing strategy across multiple MR tiers . The MR PIC code has enabled 3D simulations of laser-matter interactions on Frontier, Fugaku, and Summit, which until now were beyond the reach of standard codes. These simulations removed a major limitation of laser-based compact electron accelerators, which are promising candidates for next-generation high-energy physics experiments and ultra-high dose rate FLASH radiotherapy.

According to the SC22 schedule, this team includes researchers from Arm, Atos, CEA-Université Paris-Saclay, ENSTA Paris, GENCI, Lawrence Berkeley National Lab and RIKEN.

While the plasma accelerator team won the award, the other five nominees represent exemplary uses of some of the world’s fastest supercomputers. You can learn a lot more about each of the other finalists on this link.

The ACM also awarded its special Gordon Bell Prize for Covid-19 research based on high-performance computing for the third consecutive year. To learn more about this year’s winner of this award, Click here.