SC20 Wrap-up: From the Gordon Bell Prize to Machine Learning, Quantum Computing, and More
November 23, 2020
As they have for nearly three decades, Berkeley Lab Computing Sciences Area staff from the Computational Research Division, ESnet, and NERSC shared their expertise with the global HPC community at SC20 – albeit this year in virtual mode. From tutorials and workshops to papers, posters, panels, and more, it was a busy couple of weeks for all involved.
Here are a few highlights:
Gordon Bell Prize
CRD researchers were part of the team that won the 2020 ACM Gordon Bell Prize for DeePMD-kit, a new machine-learning-based software package that uses neural networks to enhance molecular dynamics modeling. They demonstrated what can be achieved by integrating physics-based modeling and simulation, machine learning, and efficient implementation on a next-generation computational platform.
In addition, researchers from CRD and NERSC were lead authors on a paper that was a finalist for this year’s Gordon Bell Prize. The team demonstrated how advancements to the BerkeleyGW materials science code enables large-scale, excited-state calculations to run in just minutes on HPC systems.
Best Paper Award
“The Performance and Energy Efficiency Potential of FPGAs in Scientific Computing” won the best paper award at the 11th IEEE International Workshop on Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems, held in conjunction with SC20. Authors: Tan Nguyen (CRD), Samuel Williams (CRD), Marco Siracusa, Colin MacLean (NERSC), Douglas Doerfler (NERSC), Nicholas Wright (NERSC)
Best Presentation Award
Researchers from Berkeley Lab’s Advanced Light Source, the CAMERA division, and NERSC won the Best Presentation Award at the XLOOP workshop, held in conjunction with SC20, for “Interactive Parallel Workflows for Synchrotron Tomography.” Authors: Dula Parkinson (ALS), Harinarayan Krishnan (CRD), Daniela Ushizima (CRD), Matthew Henderson (CRD), Shreyas Cholia (NERSC)
Best Research Poster Finalist
While interning at NERSC this year, University of Arizona Ph.D. Student Kevin Luna used deep learning to accelerate traditional PDE-based simulations in real time. This work earned him a spot as a finalist in the Best Research Poster competition at SC20.
Best Student Paper Finalist
A paper describing MeshfreeFlowNet — an open-source, physics-constrained, deep-learning approach for enhancing the spatial and temporal resolution of scientific data — was a finalist for the Best Student Paper Award at SC20. The lead author, Max Jiang of UC Berkeley, has been affiliated with NERSC since 2018.
1st International Workshop on Quantum Computing Software
This first-of-its-kind SC20 event explored the innovative software that is needed to make quantum computing practical and accessible. Learn more about the workshop in this Q&A with CRD’s Bert de Jong, who was on the program committee.
Deep Learning Improves High-Performance Networking
ESnet’s Mariam Kiran gave a lightning talk at SC20 on DAPHNE (Deep Learning and Artificial Intelligence for High-Performance Networks), a project she leads that is developing next-generation software tools to help scientists better predict the best time and date to schedule large-scale data transfers across ESnet.
Beyond Moore’s Law: Disaggregated Architectures
CRD’s John Shalf was part of a panel discussion on emerging high performance computing platforms at a webinar on “Disaggregated System Architectures for Next Generation HPC and AI Workloads,” held during SC20.
About Computing Sciences at Berkeley Lab
High performance computing plays a critical role in scientific discovery. Researchers increasingly rely on advances in computer science, mathematics, computational science, data science, and large-scale computing and networking to increase our understanding of ourselves, our planet, and our universe. Berkeley Lab’s Computing Sciences Area researches, develops, and deploys new foundations, tools, and technologies to meet these needs and to advance research across a broad range of scientific disciplines.