High-performance computing (HPC) has been a defining strength of Lawrence Livermore since its founding. Indeed, Livermore scientists have designed and used some of the world’s most powerful computers to drive breakthroughs in nearly every mission area. Today, the Laboratory is recognized as a world leader in the application of HPC to complex science, technology, and engineering challenges. Most importantly, HPC has been integral to the National Nuclear Security Administration’s (NNSA’s) Stockpile Stewardship Program—designed to ensure the safety, security, and reliability of our nuclear deterrent without nuclear testing. The program celebrated its 20th anniversary in 2015, and much of the credit for its overwhelming success belongs to HPC capabilities and the expertise of scientists at NNSA’s national security laboratories.
A critical factor behind Lawrence Livermore’s preeminence in HPC is the ongoing investments made by the Laboratory Directed Research and Development (LDRD) Program in cutting-edge concepts to enable efficient utilization of these powerful machines. Congress established the LDRD Program in 1991 to maintain the technical vitality of the Department of Energy (DOE) national laboratories. LDRD has been, and continues to be, an essential tool for exploring anticipated needs that lie beyond the planning horizon of our programs and for attracting the next generation of talented visionaries. Through LDRD, Livermore researchers can examine future challenges, propose and explore innovative solutions, and deliver creative approaches to support our missions. Examples of LDRD successes are numerous and encompass all Laboratory mission areas. For HPC, we developed scalable algorithms in areas including adaptive mesh refinement and algebraic multigrid solvers, created a new paradigm in statistical debugging techniques, and produced methodologies that have resulted in the largest discrete event simulations ever performed. The present scientific and technical strengths of the Laboratory are, in large part, a product of past LDRD investments.
Despite the enormous power of today’s supercomputers, we continue to demand even more capable machines to predict, with the requisite confidence, the behavior of complex physical systems. LDRD researchers are looking ahead to when the first exascale supercomputer arrives at Livermore. Exascale machines are expected to contain an estimated one billion processing elements (compared to today’s several million), making them capable of performing a quintillion (1018) floating-point operations per second.
Extreme-scale (exascale and higher) systems architecture will entail a significant departure from present-day hardware, and the methods and tools we use will have to change dramatically to exploit these systems’ capabilities. In many ways, extreme-scale computing requires that we rethink our current algorithms, simulation codes, and many other aspects of supercomputing. To the extent possible, advanced computer architectures must be co-designed with software experts. Lawrence Livermore is one of only a handful of institutions worldwide with expertise in vertical integration that enables this co-design process—from understanding and developing leading-edge hardware to implementing multiphysics applications and data-science analytics.
The feature article, Laying the Groundwork for Extreme-Scale Computing, describes how we are focusing some of our LDRD investments on addressing longer term research and development for extreme-scale computing. Six current LDRD-funded research projects are described. These projects involve ways to reduce, monitor, and manage power consumption; make systems more resilient to hardware and software faults and failures; lessen the required communication between hardware components; and create advanced algorithms that solve computing steps in parallel rather than in sequence for faster results. In addition, new algorithms are being developed for producing better quality simulations of shock hydrodynamics and first-principles molecular dynamics simulations that encompass far more atoms. Together, these efforts help Livermore simulation codes to efficiently utilize exascale and potentially other extreme-scale computers.
World-class HPC attracts new talent to Livermore and is a foundation upon which we continue to build new programs to meet national security challenges. When the Laboratory receives its first exascale machine, we are confident that our early LDRD investments in extreme-scale computing will have helped develop the requisite computational tools to ensure these machines meet their enormous potential for advancing national security and keeping Livermore at the forefront of HPC.