SC21 Recap: Berkeley Lab Computing Sciences Research Garners Attention

Staff from Berkeley Lab Computing Sciences Area once again made significant contributions to the technical program and related workshops at the annual International Conference for High Performance Computing, Networking, Storage, and Analysis (SC21), held this year in person and online Nov. 14-19. 

Best Paper

“Exploring the BBRv2 Congestion Control Algorithm for use on Data Transfer Nodes” won best paper at the 8th IEEE/ACM International Workshop on Innovating the Network for Data-Intensive Science (INDIS 2021), held in conjunction with SC21. Brian Tierney, Eli Dart, and Ezra Kissel of ESnet were co-authors on the paper, along with Eashan Adhikarla of Lehigh University. More information, including a Q&A with the authors and a link to the paper, can be found in this ESnet blog post.

Best Paper

“Requirements for Deep-learning Workloads in HPC Environments” won best paper at the 12th IEEE International Workshop on Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems (PMBS21), held in conjunction with SC21. Authors were Khaled Ibrahim, Tan Nguyen, Hai Ah Nam, Wahid Bhimji, Steven Farrell, Leonid Oliker, Michael Rowan, Nick Wright, and Samuel Williams of Berkeley Lab Computing Sciences.

Best Poster

“Enabling Combustion Simulations for Future Exascale Machines” won best research poster at SC21. The multi-lab team of authors included Berkeley Lab Computing Sciences’ Ann Almgren and Weiqun Zhang.

TOP500

Berkeley Lab’s Perlmutter supercomputer landed in the #5 spot in the latest round of the TOP500 with a performance benchmark of 70.9 Petaflop/s. Perlmutter was also #5 on the previous TOP500 list announced in June, when it reported a performance benchmark of 64.6 Petaflop/s.

The TOP500 list has now incorporated the High-Performance Conjugate Gradient (HPCG) Benchmark results, which provide an alternative metric for assessing supercomputer performance. Perlmutter took the #3 spot on the latest HPCG list with 1.91 HPCG petaflops.

Gordon Bell Finalist

“Billion atom molecular dynamics simulations of carbon at extreme conditions and experimental time and length scales.” Authors included Rahulkumar Gayatri, an application performance specialist at NERSC who is involved with the Exascale Atomistic Capability for Accuracy, Length, and Time (EXAALT) project. The EXAALT library was instrumental in supporting this research. For more information, read this article, which includes excerpts from an interview with Gayatri.

Best Paper Finalist, Computational Biology

“Accelerating Large Scale de Novo Metagenome Assembly Using GPUs.” Authors included Berkeley Lab’s Muaaz Gul Awan, Steven Hofmeyr, Rob Egan, Nan Ding, Aydin Buluc, Jack Deslippe, Leonid Oliker, and Katherine Yelick.

HPCWire Editors’/Readers’ Choice Awards

Researchers from Berkeley Lab, NERSC, JGI, and ESnet were recognized for outstanding science, technology, collaboration, and diversity/inclusion achievements in 2021 as part of the annual HPCwire Readers’ and Editors’ Choice Awards.