Standing Up the El Capitan Exascale Supercomputer

Building El Capitan, the National Nuclear Security Administration’s first exascale-class supercomputer, was an "enormous undertaking of meticulous planning and strategic collaboration,” according to a five-part article series about the project’s preparation and deployment.

"El Capitan will come online in 2024 with the processing power of more than 2 exaflops, or 2 quintillion calculations per second. The system will be used for predictive modeling and simulation in support of the stockpile stewardship program,” the first article says.

El Capitan’s design "combines HPE/Cray's advanced Slingshot interconnection network alongside processors from AMD” and will use the Advanced Simulation and Computing (ASC) program’s Tri-Lab Operating System Stack (or TOSS). (Tri-Lab refers to the three national labs: Livermore, Los Alamos, and Sandia.)

LLNL is also the first supercomputing center to use HPE’s near-node local storage solution, called Rabbits. "El Capitan will have the global Lustre file system that’s shared between all the nodes as well as HPE’s dynamically configurable Rabbits,” says Livermore Computing computer scientist Brian Behlendorf, who also serves as I/O lead for El Capitan.

Read more at Computing at LLNL.