Lawrence Livermore National Laboratory (LLNL) has deployed another new supercomputer: Ruby. The lab, which works under the umbrella of the National Nuclear Security Administration (NNSA), says the new cluster will serve research tasks relating to nuclear security and COVID-19 research.
Ruby, provided by Supermicro, is already operational and consists of 1,512 nodes: 1,480 batch nodes equipped with Intel Xeon CLX-8276L CPUs, 24 debug nodes, and eight login nodes, all networked with Cornelis Networks Omni-Path and liquid-cooled using a direct-to-chip approach. In total, the system has 290,304 GB of memory and reports 5,959.2 peak teraflops of computing power. While LLNL did not report a Linpack score, the lab expects that Ruby will rank within the top 100 of the next Top500 list, which will be announced in just a few days at SC20.
Much of this computer power is being put to work on nuclear science research under the purview of the NNSA. The system is already simulating plasma dynamics, neutron production and more for various groups at LLNL.
“Ruby provides a substantial computing resource in our open collaboration zone, which has experienced a heavy increase in demand due to an uptick in telecommuting and a growth in external collaborations,” said Chris Clouse, acting program director for LLNL’s Advanced Simulation and Computing (ASC) program. “A resource like Ruby provides a venue for leveraging expertise and tools in the open community for areas that are important to our programmatic missions.”
In addition to nuclear science research, LLNL researchers are also already using Ruby for COVID-19 research – namely, drug discovery-oriented molecular docking calculations aimed at identifying small molecules that can bind to SARS-CoV-2’s proteins. Indeed, Ruby was partially funded through the Coronavirus Aid, Relief and Economic Security (CARES) Act, which has been disbursing funds to Department of Energy (DOE) labs for COVID-19 research and associated infrastructure since late March.
“Ruby is excellent for running the molecular docking calculations,” said Felice Lightstone, leader of LLNL’s Biochemical and Biophysical Systems Group and head of the COVID-19 small molecule work. “Our early access on Ruby is allowing us to screen about 130 million compounds per day when using the entire machine. As our COVID-19 therapeutic effort moves toward optimizing compounds we have identified as promising, Ruby allows us to maximize the throughput of our new designs.”
The Ruby announcement comes hot on the heels of LLNL’s announcement of the high-memory Mammoth cluster, which is being purpose-delivered with COVID-19 research in mind. Mammoth, which is already hard at work on COVID-19 research projects, delivers around 294 peak teraflops of computing power. Like Ruby, Mammoth received funding from the CARES Act.
The announcement also comes on the heels of some other news: the unexpected resignation of Lisa E. Gordon-Hagerty, the NNSA’s administrator, in the immediate wake of Donald Trump’s loss to Joe Biden in the U.S. presidential election. William Bookless, formerly the NNSA’s principal deputy administrator, is now serving as acting NNSA administrator.
Header image: the Ruby supercomputer. Image courtesy of Katrina Trujillo/LLNL.