NERSC

NERSC
Name	National Energy Research Scientific Computing Center
Established	1974
Research field	High-performance computing, Computational science
Location	Lawrence Berkeley National Laboratory, Berkeley, California
Operating agency	United States Department of Energy

NERSC. The National Energy Research Scientific Computing Center is the primary high-performance computing facility for scientific research sponsored by the Office of Science within the United States Department of Energy. Located at Lawrence Berkeley National Laboratory in Berkeley, California, it provides computational resources and expertise to thousands of researchers nationwide, supporting a vast portfolio of work in fields like climate science, materials science, high-energy physics, and computational biology. As one of the largest facilities of its kind in the world, it operates some of the most powerful supercomputers dedicated to open scientific research, enabling breakthroughs across the DOE mission space.

Overview

The center serves as the mission high-performance computing facility for the DOE Office of Science, providing essential computational infrastructure to a broad community of researchers. Its primary goal is to accelerate the pace of scientific discovery by deploying and supporting leading-edge computing systems and storage systems for computationally intensive projects. NERSC supports a highly diverse workload, from large-scale simulations of the universe to detailed modeling of novel energy materials, requiring robust software environments and expert staff. The facility operates under a long-standing partnership between the DOE and Lawrence Berkeley National Laboratory, leveraging the laboratory's deep expertise in computational science and applied mathematics.

History

The origins of the center trace back to 1974 when it was established as the Controlled Thermonuclear Research Computer Center to support fusion energy research at Lawrence Livermore National Laboratory. In 1990, the facility was relocated to Lawrence Berkeley National Laboratory and renamed the National Energy Research Supercomputer Center, broadening its mission to serve the entire DOE Office of Science research community. A significant milestone was reached in 1996 with the deployment of the SP2 system, a massively parallel IBM supercomputer that marked a shift towards new architectures. The center has continuously evolved, deploying pioneering systems like Cray T3E, IBM POWER, and Cray XT4 machines, each advancing the frontiers of scientific computing capability for its users.

Computing Systems

The facility has a long history of deploying and operating cutting-edge supercomputer architectures, each given a name inspired by notable scientists. Recent leadership-class systems include Cori, a Cray XC40 system, and Perlmutter, a HPE Cray machine that incorporates both CPU and GPU nodes to accelerate a wide range of scientific applications. These systems are integrated with high-performance Lustre file systems and HPSS archival storage to manage the immense data generated by simulations and experiments. The center also provides specialized testbed systems, such as those featuring ARM architecture processors or novel memory technologies, for evaluating future computing paradigms in collaboration with vendors like Intel and NVIDIA.

Scientific Impact

Research enabled by the center's resources has led to transformative advances across multiple scientific domains, resulting in numerous accolades including Gordon Bell Prize awards. In cosmology, large-scale simulations have mapped the evolution of the universe and informed observations by projects like the Dark Energy Spectroscopic Instrument. In materials science, researchers have discovered new catalyst materials and battery components critical for renewable energy technologies. The facility's computing power has been pivotal for the ATLAS experiment and CMS experiment at CERN, processing vast datasets from the Large Hadron Collider. Furthermore, climate scientists have used its systems to run high-resolution models of the Earth's atmosphere and oceans, improving predictions related to climate change.

User Programs and Allocations

Access to computational resources is primarily granted through a competitive peer-review process managed by the DOE Office of Science's Advanced Scientific Computing Research program. The primary allocation program is the INCITE program, which awards large time allocations to ambitious computational campaigns, while the ALCC program supports more focused, mission-critical research. The center also manages the NERSC Exascale Science Applications Program, which prepares key applications for next-generation exascale computing systems. User support is extensive, including training workshops, consultation on code optimization, and assistance with data analysis and visualization through tools developed in collaboration with the Scientific Data Management group.

Facility and Operations

The computing systems are housed in a state-of-the-art machine room within the Shyh Wang Hall at Lawrence Berkeley National Laboratory, designed with a focus on energy efficiency and advanced cooling technologies. Operations are supported by a dedicated staff of systems engineers, network specialists, and computational scientists who ensure high system availability and performance. The facility maintains a high-speed research network, connecting to national infrastructures like ESnet, which is also managed by Lawrence Berkeley National Laboratory, to facilitate rapid data movement for collaborators at institutions like SLAC National Accelerator Laboratory and Argonne National Laboratory. Continuous monitoring and proactive maintenance are hallmarks of the center's approach to delivering a reliable production environment for the scientific community.

Category:Supercomputer sites Category:Lawrence Berkeley National Laboratory Category:United States Department of Energy national laboratories Category:High-performance computing