The National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory (Berkeley Lab) today formally unveiled the first phase of its next-generation supercomputer, called Perlmutter. The new system will greatly increase the high-performance computing (HPC) capability for a broad spectrum of unclassified scientific research within the U.S. Department of Energy (DOE) Office of Science.
Perlmutter features a heterogeneous architecture that will provide four times the computational power currently available at NERSC, making it among the fastest supercomputers in the world for scientific simulation, data analysis, and artificial intelligence applications. It will enable a larger range of applications than previous NERSC systems and is the first NERSC supercomputer designed from the very beginning to meet the needs of both simulation and data analysis.
The Perlmutter system will play a key role in advancing scientific research in the U.S. and is front and centre in several critical technologies, including advanced computing, artificial intelligence, and data science. The system will also be heavily used in studies of the climate and the environment, clean energy technologies, semiconductors and microelectronics, and quantum information science.
The Perlmutter supercomputer will help inspire the next generation of scientists and innovators, allowing the U.S. and DOE to remain a leader in using scientific computation to answer the greatest questions. As the researchers continue to enhance and deploy computing platforms like this, their national labs will only be better positioned to develop solutions to today’s toughest problems, from climate change to cybersecurity.
To ensure that Perlmutter’s users can readily utilise this new technology, NERSC has been working with key application development teams since 2019 to prepare codes for Perlmutter through its NERSC Exascale Science Applications Program.
The system, an HPE Cray EX supercomputer, is being delivered in two phases. Phase 1 features 1,536 GPU-accelerated nodes, each containing four NVIDIA NVlink-connected A100 Tensor Core GPUs and one 3rd Gen AMD EPYC™ processor. Phase 1 also includes a 35 PB all-flash Lustre file system that will provide very high-bandwidth storage. Phase 2, set to arrive later this year, will add 3,072 CPU-only nodes, each with two 3rd Gen AMD EPYC™ processors and 512 GB of memory per node.
The launch showcases the strong collaboration between HPE and Berkeley Lab by being one of the first systems to be powered by the HPE Cray EX supercomputer, which leverages comprehensive, next-generation supercomputing technologies. They want Perlmutter to play an integral role in augmenting research efforts to support NERSC’s ongoing mission in advancing insights for developing new energy sources.
NVIDIA states that Perlmutter is a world-class supercomputer with AI capabilities that enable the scientific community to push the boundaries of supercomputing research as they enter the exascale AI era. Similarly, AMD says that there has never been a more exciting time in high-performance computing. As an industry, they are driving toward the exascale era with the most powerful supercomputers that researchers, scientists and technology leaders have ever seen.
Berkeley Lab and NERSC have a long tradition of supporting team science, which is also a core tenet of Dr Perlmutter’s legacy as a scientist, teacher, and mentor. During the unveiling event, a panel of scientists who have long used NERSC resources in their research discussed the impact that Perlmutter is expected to have on scientific discovery going forward.
Science has developed the ability to collect very large amounts of data and bring them all at one time. This allows the combination between science and the power of supercomputer facilities. The new supercomputer is exactly what scientists need to handle these datasets. As a result, they are expecting to find new discoveries in cosmology, microbiology, genetics, climate change, material sciences and almost any other field.