Specifically, the award will spotlight innovative parallel computing contributions based on performance and innovation in their computational methods and their contributions toward improving climate modeling and understanding the Earth’s climate system. 

Today, climate modeling provides the most viable window to anticipate, and potentially reduce, the future impacts of climate change driven by increasing atmospheric greenhouse gas levels. By recognizing the scientific and computing contributions required to tackle worldwide climate change and respond to the need for more accurate, rapidly available climate information, the Gordon Bell Prize for Climate Modelling can be a means to elevate awareness about the consequences of climate change while showcasing the pioneering science and dedicated researchers working on this Grand Challenge problem. 

John Taylor

ACM Gordon Bell Prize for Climate Modelling Committee Chair

John on LinkedIN

Finalists for the award are determined by the work’s actual or potential influence on the realm of climate modeling, its interconnected disciplines, and broader societal impact. The use of HPC in the climate modeling applications is also considered. Based on these criteria, the Association for Computing Machinery (ACM) GBP for Climate Modelling Award Committee selected three finalists for the inaugural award.

Target Surface Temperatures

Predicted Surface Temperatures

Finalist Project Summaries

Finalist 1

This work introduces an efficient and performance portable implementation of the Simple Cloud Resolving E3SM Atmosphere Model (SCREAM). SCREAM is a full-featured atmospheric global circulation cloud-resolving model. A significant advancement is SCREAM was developed anew using C++ and incorporates the Kokkos library to abstract the on-node execution model for both CPUs and GPUs. To date, only a few global atmosphere models have been ported to GPUs. SCREAM was able to run on both AMD and NVIDIA GPUs and on nearly an entire exascale system (Frontier). On the Frontier system, it achieved a groundbreaking performance, simulating 1.26 years per day for a practical cloud-resolving simulation. This constitutes a pivotal stride in climate modeling, offering enhanced and highly necessary predictions regarding the potential outcomes of future climate changes.

Finalist 2

This work presents a comprehensive simulation that captures the entire dynamic sequence of the Tonga eruption, spanning from the initial shock waves, earthquakes, and tsunamis to the subsequent development of mushroom clouds over the following 6-7 days. This simulation accounts for the dispersion and diffusion of ash and water vapor during the period. The modeling system was able to resolve the effects of the full coupling of the volcanic eruption, resulting seismic activity, and concurrent oceanic and atmospheric phenomena. In their paper, the team describes a novel modeling system designed for volcanic eruptions and atmosphere circulation. This system was run on a new Sunway supercomputer in China using a spatial resolution, ranging from 10 meters at local scale to 3 kilometers globally. The method incorporates an enhanced multimedium and multiphase smoothed particle hydrodynamics (SPH) model combined with a fully coupled meteorology-chemistry global atmospheric modeling scheme. The modeling system was able to use 400 billion particles with 80% parallel efficiency using 39 million processor cores. This marks a breakthrough in the ability to study interactions between tectonic processes and climate change. Furthermore, it provides the means to develop an early-warning simulation system capable of addressing similar global hazard events in the future.

Finalist 3

The work presents a real-time 30-second-refresh numerical weather prediction (NWP), performed via exclusive use of 11,580 nodes (approximately 7% of the total) of the Fugaku supercomputer during the 2021 Tokyo Olympics and Paralympics. A total of 75,248 forecasts were disseminated in the one-month period with time-to-solution in less than three minutes for each 30-minute forecast. Japan’s Big Data Assimilation (BDA) project developed the novel NWP system for precise prediction of hazardous rains as a step toward solving the global climate crisis, where hazardous rain events are likely to increase with global warming. To achieve the required time-to-solution for a real-time 30-second refresh with high accuracy, the core BDA software incorporates single precision and enhanced parallel input/output (I/O) with tailored configurations involving 1,000 ensemble members of a 500-meter resolution weather model. Compared to the conventional one-hour refresh systems used by the weather bureaus, the BDA system not only showcased a two-orders-of magnitude increase in problem size, but it also revealed the effectiveness of 30-second refresh for highly nonlinear and rapidly evolving convective rainstorms. This endeavor stands as a testament to the value of engaging advanced computational methodologies to advance understanding of intricate meteorological phenomena.

More on Gordon Bell Awards

This year’s SC23 Test of Time Award recognizes a paper presented during SC09, held in Portland. Ore., “Millisecond-Scale Molecular Dynamics Simulations on Anton.”

Reflecting the effort involved in building such a complete computing system, the paper included 22 authors: David E. Shaw, Ron O. Dror, John K. Salmon, J. P. Grossman, Kenneth M. Mackenzie, Joseph A. Bank, Cliff Young, Martin M. Deneroff, Brannon Batson, Kevin J. Bowers, Edmond Chow, Michael P. Eastwood, Douglas J. Ierardi, John L. Klepeis, Jeffrey S. Kuskin, Richard H. Larson, Kresten Lindorff-Larsen, Paul Maragakis, Mark A. Moraes, Stefano Piana, Yibing Shan, and Brian Towles.

A Historic Paper

The paper described Anton, a special-purpose computing system designed and operated by D.E. Shaw Research in New York City to accelerate molecular dynamics (MD) simulations. The machine architecture and algorithms for MD were co-designed to achieve high performance on simulations where parallelism was limited by the number of molecules and high speed was required to simulate behavior over a sufficient interval.

With 512 processors, Anton demonstrated the power of hardware-software co-design to advance complex protein research. Its processors projected the movements of 50,000 atoms over time, anticipating the actions of complex molecular relationships in real time. In short, back in 2009, Anton could do in a day what it then took the world’s top supercomputers months to achieve.

In 2010, the Pittsburgh Supercomputing Center (PSC) received a grant from the National Institutes of Health and hosted an Anton machine. A long list of university and pharmaceutical researchers were excited to have access to a resource that enabled access to atomic-level details that helped inform their work in therapeutic drug design. Now, PSC is running Anton 2 for important MD simulations.

“Since 2010, PSC has provided the national research community with unique access to the remarkable Anton supercomputers,” said PSC’s Scientific Director Dr. Philip Blood. “The game-changing performance of Anton systems—orders of magnitude faster than any other resource for biomolecular simulation—continues to enable breakthrough scientific discoveries that would not have been possible through any other means.”

“The game-changing performance of Anton systems—orders of magnitude faster than any other resource for biomolecular simulation—continues to enable breakthrough scientific discoveries that would not have been possible through any other means.”

— Dr. Philip Blood, Scientific Director, PSC

Beyond its influence on MD research, the Anton system exemplified what has become increasingly common: end-to-end co-design of architecture and hardware, system software, and algorithms to maximize delivered performance. With the end of Moore’s law and increasing limits to commodity hardware performance, Anton was a pathfinder for the continued evolution of high-performance computing.

Today, D.E. Shaw Research continues to develop Anton systems primarily used for internal research. However, there is an Anton system available for non-commercial research use by U.S. universities and other not-for-profit institutions.

“We feel honored to receive this award,” said Dr. David E. Shaw, chief scientist at D.E. Shaw Research and first author of the original paper released during SC09. “We regard this as an acknowledgment not only of our own work, but of the scientific contributions made by 239 research groups using Anton—time we’ve made available without cost to universities and other nonprofit institutions.”

Read More:

Shaw DE, RO Dror, JK Salmon, et al. 2009. “Millisecond-scale molecular dynamics simulations on Anton,” In Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis (SC09), Portland, OR, USA, pp. 1-11, doi: 10.1145/1654059.1654126.

The GBP is awarded to the most valuable scientific computation as demonstrated using state-of-the-art software and hardware technologies on world-leading supercomputers. The GBP represents various aspects of evaluation, such as the importance of target problems, performance optimization, maximum utilization of target system performance, and knowledge given for widely spread platforms. This year, the Gordon Bell finalists’ work spans various applications, including materials science, fluid dynamics, nuclear simulation, seismic processing, and biomolecular simulations. The hardware platforms also include world-class systems: Frontier (ORNL, USA), new Sunway System (SC Wuxi, China), LUMI (EuroHPC/CSC, Finland), Leonardo (EuroHPC/Cineca, Italy), Cerebras CS-2 (KAUST, Saudi Arabia), and Perlmutter (NERSC, USA).

The following summarizes this year’s Gordon Bell finalists via brief descriptions of the work. Please note that the results or system sizes will not be finalized before the teams’ final submissions in August. The ACM Gordon Bell Prize award winner will be announced as part of a special ceremony at SC23 this November in Denver.

Taisuke Boku

ACM Gordon Bell Prize 2023 Committee Chair; Professor/Director of the Center for Computational Sciences, University of Tsukuba

Finalist 1

In this work, the team developed a mixed method to combine density function theory (DFT) and the quantum many body (QMB) problem using a machine learning technique. The effort achieves high accuracy of calculation and affords large-scale modeling with the inverse-DFT that links the QMB method to DFT. They realized the ground-stage energy calculation while keeping the accuracy commensurate with QMB, using more than 60% of resources on the Frontier supercomputer housed within the Oak Ridge Leadership Computing Facility.

Finalist 2

The team developed a new large eddy simulation code to solve compressible flows in turbomachinery. They applied it to NASA’s grand challenge problem with a high-order unstructured solver for a high-pressure turbine cascade of 1.69 billion mesh elements and 865 billion degrees of freedom. The code has been calculated on Wuxi’s new Sunway supercomputer with extreme many-cores per node, up to 19.2 million cores, where each computation node consists of 384 calculation cores and six control cores.

Finalist 3

The team simulated an advanced nuclear reactor system coupling radiation transport with heat and fluid simulation, including the high-fidelity, high-resolution Monte-Carlo code, Shift, and the computational fluid dynamics code, NekRS. Nek5000/RS was implemented on ORNL’s Frontier system and achieved 1 billion spectral elements and 350 billion degrees of freedom, while Shift achieved very high weak-scaling on 8192 system nodes. As a result, they calculated six reactions in 214,896 fuel pin regions below 1% statistical error, yielding first-of-a-kind resolution for a Monte Carlo transport application.

Finalist 4

The team developed a high-fidelity spectral-element code, Neko, which is essential for unprecedented large-scale direct numerical simulations of fully developed turbulence—all while maintaining high-performance portability on GPU-accelerated platforms. They applied a GPU-optimized preconditioner with task overlapping for the pressure-Poisson equation and in situ data compression. They also conducted initial runs of Rayleigh–Bénard convection at extreme scale on the LUMI and Leonardo supercomputers with up to 16,384 GPUs via a sophisticated workflow control.

Finalist 5

This work exploits the high-memory bandwidth of artificial intelligence (AI)-customized Cerebras CS-2 systems for seismic processing by leveraging the low-rank matrix approximation to fit the problem on SRAM (static random-access memory) wafer-scale hardware and use many wave-equation-based algorithms that rely on multidimensional convolution operators. As a result, the team implemented a standard seismic benchmark dataset into the small local memories of Cerebras processing elements, extrapolating a worst-case load-balanced whole application execution to 48 CS-2 systems on 35,784,000 processing elements. This is a significant example of applications run on AI-customized architectures that can enable a new generation of seismic algorithms.

Finalist 6

The group developed the Allegro architecture to bridge the accuracy-speed tradeoff of atomistic simulations and enable the description of dynamics in structures of unprecedented complexity at quantum fidelity. This is achieved through a combination of innovative model architecture, massive parallelization, and model implementations optimized for efficient GPU utilization. Allegro’s scalability is illustrated by a nanoseconds-long stable simulation of protein dynamics and up to 44-million atom structure of a complete, all-atom, explicitly solvated HIV capsid on the Perlmutter system at the National Energy Research Scientific Computing Center. They achieved strong scaling up to 100 million atoms.

More on Gordon Bell Awards

Nominations are being accepted until June 30, 2023.

The most successful nominations draw attention to the candidate’s impact on the field of HPC both in terms of technical contributions to the state of the industry, as well as service and mentoring activity. Past recipients of the award stand as outstanding examples of success that have helped shape the future of HPC, and cultivated a more diverse, inclusive, and innovative environment. 

If you know any women who fit the description above, we encourage you to nominate them and recognize the work done by these current and future leaders of the industry.

Learn More & Nominate

Make a Nomination

We invite you to take part in this celebration of excellence by nominating exceptional candidates who have made a significant impact in the HPC community. Discover all the necessary details about the various Society Awards, the nomination process, and the criteria for selection.

These awards aim to honor the incredible contributions made by researchers, developers, and professionals who are harnessing the power of HPC to drive groundbreaking advancements and innovations in various sectors, such as science, engineering, healthcare, and more. Through their dedication, expertise, and passion, these trailblazers have demonstrated excellence in utilizing supercomputing resources to address complex challenges, optimize processes, and unlock new frontiers of knowledge.

From the first award recognizing William Pugh’s work that originated the Omega Test to last year’s acknowledgement of Chung-Hsing Hsu and Wu-chun Feng’s power-aware algorithm that extended the notion of low-power to HPC systems, each Test of Time Award has spotlighted among the most discussed and deliberated achievements and innovators in HPC. Notably, the award affords a distinct opportunity to celebrate the ideas that have redefined how to manage, build, and evolve HPC.

The Test of Time Award also is a reminder that the SC Conference itself is one of the greatest resources for those seeking to make a lasting impression in computing research and development. Partially, the award was initiated as an incentive to encourage up-and-coming researchers and students to bring their best work to the conference for a chance to leave their own indelible mark on HPC history.

Nominations for SC23’s Test of Time Award will be accepted starting on Friday, March 10, 2023. The process is straightforward, and any paper published as part of SC Conference Proceedings between 10 and 25 years ago is eligible.

As SC23 marks the conference’s 35th year, surely there are many deserving HPC moments that readily stand the test of time.

Learn More & Nominate