NERSC’s next flagship supercomputer, named in honor of Nobelist Jennifer Doudna, will feature two different storage systems to meet the broad and growing needs of science. - Credit: Margie Wylie, Berkeley Lab (This work is a derivative of “Jennifer Doudna by Christopher Michel in 2023,” used under CC BY 4.0 via Wikimedia Commons.)
The upcoming Doudna supercomputer at the National Energy Research Scientific Computing Center (NERSC) will partner next-generation high performance computing (HPC) capabilities with cutting-edge data storage solutions to meet the rapidly evolving needs of science users. With technology provided by VAST Data and IBM, Doudna will feature storage performance up to five times faster than NERSC’s current system and offer performance guarantees for time-sensitive science.
The new system must be able to handle the established large-scale workloads of research areas like molecular dynamics and geophysical modeling but also meet the data-driven demands of expanding workloads like AI training and inference.
“Scientific workloads are evolving into complex workflows to leverage the new opportunities from integrating simulation and modeling, AI, and data growth,” said Hai Ah Nam, NERSC-10 Project Director. “HPC system design must also evolve and address these changing requirements head-on, especially with regards to how we manage and use data. With Doudna, we’re tackling critical bottlenecks in the scientific workflow to accelerate scientific discovery.”
To ensure consistent, predictable performance for workflows that require near-real-time data analysis and maximum performance for applications that need it, Doudna will have two storage systems: a quality-of-service storage system (QSS) and a platform storage system (PSS).
The QSS will be powered by VAST Data, whose forward-looking AI Operating System unifies data storage, database, compute, messaging, and reasoning capabilities into a single, data-centric infrastructure built from the ground up for AI and agentic workflows. The system will allow NERSC to provide deadline-dependent workloads with performance guarantees through either persistent or schedulable performance. This seamless service will particularly benefit science teams whose workflows call for time-constrained analysis, such as those taking readings and gathering data via telescopes, particle accelerators, and other experimental infrastructure.
“The future of supercomputing is delivering cloud-like simplicity and control at the scale and intensity of the world’s largest research environments,” said Jeff Denworth, co-founder of VAST Data. “With the VAST AI Operating System, NERSC is pioneering a new model for Doudna, where users get guaranteed performance, security, real-time access, and built-in data services – without the operational friction of traditional HPC systems. Together, NERSC and VAST are setting the blueprint for exascale computing, enabling breakthrough capabilities that will define the next era of scientific computation.”
The PSS will be provided by IBM and function as an all-flash high-performance parallel scratch filesystem based on IBM’s software-defined file and object storage platform, IBM Storage Scale. Engineered for modern, data-intensive HPC-driven research, AI, and advanced analytics, IBM Storage Scale delivers high speed, scalable performance, and automated efficiency designed to help eliminate bottlenecks and streamline data workflows, empowering researchers to focus on discovery instead of infrastructure management.
"As America’s scientists lead the charge in fields like AI, quantum computing, and data-driven research, systems like the Doudna supercomputer demand storage solutions that are not only scalable and high-performing, but also intelligent, resilient, and easy to manage,” said Vanessa Hunt, General Manager, Technology, US Federal Market for IBM. “IBM Storage Scale is purpose-built to support the next wave of American innovation – delivering the speed, flexibility, and reliability needed to power breakthrough discoveries, while simplifying data management in even the most complex HPC environments.”
NERSC has remained at the forefront of innovation due to its agility and mission to meet the needs of the scientific community. The storage solutions for Doudna have grown from the diverse needs of NERSC’s users. Though Doudna’s storage will need to accommodate a tenfold increase in computational power over the current Perlmutter system, more than brute force will be required to simultaneously satisfy requirements across the wide spectrum of the most disparate workloads. The flexibility offered by Doudna’s innovative storage solutions will offer researchers unprecedented precision for complex science.
“Doudna’s storage solutions will leverage new technology and techniques and build on the high-performance solid-state solutions NERSC has helped evolve,” said NERSC HPC architecture and performance engineer Stephen Simms. “The addition of quality-of-service will provide predictable performance through fine-grained control of file system capability. This partnership will further our aim to enhance the user experience in the service of science.”
The National Energy Research Scientific Computing Center (NERSC) is the mission computing facility for the U.S. Department of Energy Office of Science, the nation’s single largest supporter of basic research in the physical sciences.
Located at Lawrence Berkeley National Laboratory (Berkeley Lab), NERSC serves 11,000 scientists at national laboratories and universities researching a wide range of problems in climate, fusion energy, materials sciences, physics, chemistry, computational biology, and other disciplines. An average of 2,000 peer-reviewed science results a year rely on NERSC resources and expertise, which has also supported the work of seven Nobel Prize-winning scientists and teams.
NERSC is a U.S. Department of Energy Office of Science User Facility.