Agencies lean into supercomputing
High-performance systems will support research to maintain the U.S. nuclear weapons stockpile, improve weather forecasting for military operations and deliver containerized supercomputing at the tactical edge.
The Department of Energy’s National Nuclear Security Administration is getting its first exascale supercomputer to help run the mission-critical research required to maintain the U.S. nuclear weapons stockpile.
Built by Cray, "El Capitan" will be delivered in late 2022 and housed at Lawrence Livermore National Laboratory (LLNL) for the Stockpile Stewardship Program, which ensures the safety, security and effectiveness of the nation’s nuclear stockpile in the absence of underground testing.
The Cray system will allow researchers from all of NNSA’s weapons design laboratories -- LLNL, Los Alamos National Laboratory and Sandia National Laboratories -- to run GPU-powered 3D simulations and calculations at resolutions that today’s supercomputers can't practically provide. Cray’s new Shasta architecture will allow El Capitan to process nuclear security applications at more than 1.5 exaflops -- more than 50 times the speed of LLNL's Sequoia system and substantially outpacing its Sierra system, which is currently the world’s second-most powerful supercomputer at 125 petaflops of peak performance.
El Capitan will be DOE’s third exascale-class supercomputer, following Argonne National Laboratory’s Aurora and Oak Ridge National Laboratory’s Frontier system. All three DOE exascale supercomputers will be built by Cray and feature the Shasta architecture, Slingshot interconnect and software platform.
El Capitan will be used to make critical assessments necessary for addressing evolving threats to national security and other issues such as nonproliferation, nuclear counterterrorism and nuclear weapon aging issues for which researchers have no nuclear test data.
Defense supercomputing
The Defense Department's High Performance Computing Modernization Program (HPCMP) announced that it had selected two Cray CS500 systems for its TI-18 annual technology procurement.
The CS500 is a cluster supercomputer designed for data-intensive workloads. It features industry-standard server nodes and components have been optimized for high-performance computing and tightly integrated with a comprehensive HPC software stack.
The Army Research Lab and the Army Engineering and Research Development Center will each deploy a Cray CS500 to help serve the U.S. through accelerated research in science and technology.
The CS500 systems, which will be delivered in the fourth quarter of 2019, will be equipped with 2nd Gen AMD EPYC processors and NVIDIA Tensor Core GPUs, enabling new discoveries across diverse research disciplines to address the Department’s most challenging problems.
In June, HPCMP announced a portable IBM supercomputer for the DOD Test and Evaluation and the Acquisition Engineering, and Science and Technology communities.
The new supercomputer will initially be based at DOD Supercomputing Resource Center and will serve users from all of the services and agencies of the department. This "HPC in a container" is designed to be deployable to the tactical edge, HPCMP officials said. Deployment opportunities to remote locations are currently being explored and evaluated.
The system will be housed in a shipping container with onboard power conditioning and cooling, along with the corresponding hardware and software maintenance services. The system will provide over 6 petaflops of single precision performance for training and inference machine learning workloads. It also has over 1 petabyte of high-performance, solid-state storage for data analytics workloads, HPCMP said. The system will support military use cases that were not possible with supercomputers installed in fixed facilities.
The system is expected to be delivered later in 2019 and enter production service shortly after.
The Air Force Life Cycle Management Center in partnership with Oak Ridge National Laboratory (ORNL) announced acquisition of a Cray Shasta supercomputer for operational weather forecasting and meteorology. The new system, named HPC11, will enable higher fidelity weather forecasts for military operations worldwide.
ORNL will provide supercomputing-as-a-service on the HPC11 Shasta system to the Air Force 557th Weather Wing, which develops and delivers comprehensive terrestrial and space weather information to the Air Force and Army. The new system will feature the Cray Slingshot interconnect, which will boost performance of time-critical weather prediction workloads and enhance the Air Force’s ability to create improved weather forecasts and weather threat assessments ahead of missions. The system is expected to be delivered in late 2019.
NEXT STORY: Why Speed Kills Cities