CORAL
CORAL was a first-of-its-kind U.S. Department of Energy (DOE) collaboration between the National Nuclear Security Administration’s (NNSA’s) ASC Program and the Office of Science’s Advanced Scientific Computing Research program (ASCR) that culminated in three ultra-high performance supercomputers at Lawrence Livermore, Oak Ridge, and Argonne national laboratories. The systems, delivered in the 2017 timeframe, were used for the most demanding scientific and national security simulation and modeling applications, and enabled continued U.S. leadership in computing. The Livermore system resulting from CORAL was named Sierra.
CORAL was the next major phase in the U.S. Department of Energy’s scientific computing roadmap and path to exascale computing. The procurements resulting from CORAL have influenced the modernization of future generations of computing throughout the NNSA complex.
Sierra
Sierra was the preceding advanced technology high performance computing (HPC) system to El Capitan for NNSA’s ASC Program, and was sited at Lawrence Livermore in 2017. Purchased through the CORAL request for proposal process, Sierra provided over five times the sustained scalable science performance of its predecessor, Sequoia.
Sierra provided computational resources that were essential for nuclear weapon scientists to fulfill the stockpile stewardship mission through simulation in lieu of underground testing. Modern simulations on powerful computing systems are key to supporting the annual assessment of all stockpile systems, resolving significant finding investigations (SFIs), accomplishing upcoming life extension programs (LEPs) with computationally taxing advanced safety and surety features, and for supporting qualification of hostile environments, safety calculations of abnormal environments, and gravity and reentry simulations.
Sierra’s architecture
Sierra provided over five times Sequoia's workload performance with a 125 petaflop/s peak. The Sierra system included compute nodes (POWER9 Architecture Processor, NVIDIA Volta GPUs, NVMe-comaptible PCIe 800 GB SSD, greater than 512 GB DDR4 + HBM, and coherent shared memory) and compute racks (standard 19-inch with warm-water cooling). The compute system included 1.38 petabytes of memory, provided 125 petaflop/s peak performance, and used 11 MW of power. The Global Parallel File System had 120 PB usable storage and 1.0 TB/s bandwidth.
How was Sierra used?
Sierra served the NNSA’s ASC program to ensure the safety, security, and effectiveness of the nation’s nuclear deterrent without testing. Simulation is the integrating element of the science and engineering-based program to maintain the nation’s aging nuclear stockpile. Scientists and engineers in NNSA’s Advanced Simulation and Computing (ASC) program used Sierra to assess the performance of integrated weapon systems as well as for science and engineering calculations. Key scientific areas include: materials modeling problems; turbulent flow and instabilities in such flows; and so-called laser plasma calculations, necessary to assess and predict the performance of the National Ignition Facility (NIF) in its support of the program. This work has important implications for other national security concerns, including nonproliferation and counterterrorism. The new capabilities developed on Sierra were are being applied to the study of other global and national challenges such as energy, earth systems modeling, and medicine.
Where can I find more information?
- Livermore joins Oak Ridge and Argonne to develop next supercomputers
- Argonne Leadership Computing Facility
- Oak Ridge Leadership Computing Facility
- High Performance Computing at Lawrence Livermore
- Advanced Simulation and Computing Program at Lawrence Livermore
- National Advanced Simulation and Computing Program
- Multiprogrammatic and Institutional Computing at Lawrence Livermore
