CADRE Symposium - Keeping the World Safe with Extreme Scale Data Management

Gary Grider is the leader of the High Performance Computing (HPC) Division at Los Alamos National Laboratory. Los Alamos’ HPC Division operates one of the largest supercomputing centers in the world focused on US National Security for the US/DOE National Nuclear Security Administration. As Division Leader, Gary is responsible for all aspects of High Performance Computing research, technologies development, and deployment at Los Alamos. Gary has 30 granted patents and over 13 pending in the data storage area and has been working in HPC and HPC related storage since 1984.

Los Alamos has had a long history in HPC from the 1940s to present. The 1992 nuclear test ban led to the need to do full scale nuclear weapons testing using digital simulation. These virtual tests involve extreme scale, multi-resolution, multi-physics, and multi-link scale simulations that run on a million cores with a petabyte of in memory working data. These simulations routinely run for as much as a year to produce a useful test result. This type of computing can generate hundreds of petabytes of data to be managed in weeks to months. Managing the data from irreproducible nuclear tests, sub critical materials tests, and simulations consisting of a billion files and exabytes of data for 70 years has led Los Alamos to be near the cutting edge of at scale data management methods. This presentation will cover the background of why and how we simulate and concentrate on the needed data management tools that have been developed or are in development to enable scientists to gain scientific insight to use in decision support for US nuclear stockpile stewardship.