Samuel Williams Honored with E.O. Lawrence Award

Samuel Williams, a senior scientist in the Lawrence Berkeley National Laboratory’s (Berkeley Lab’s) Applied Mathematics and Computational Research Division (AMCR) and the National Energy Research Scientific Computing Center (NERSC) is the recipient of the U.S. Department of Energy’s (DOE’s) Ernest Orlando Lawrence Award in the category of Computer, Information, and Knowledge Sciences. Williams is being recognized for “pioneering contributions in performance modeling, hardware/software design optimization, and novel computer architecture for high-performance computing (HPC), enabling the modeling of complex computational solutions and increasing analysis speed, simulation accuracy, and computational efficiency.” 

The laureates will be celebrated at a ceremony on November 13, 2026, in Washington, D.C. Established in 1959 to honor Berkeley Lab founder and Nobel Laureate Ernest Orlando Lawrence, who ushered in the era of multidisciplinary big science, the award recognizes mid-career U.S. scientists for exceptional contributions supporting the DOE’s mission. 

“This year’s Lawrence Award recipients exemplify the ingenuity and determination that drive America’s leadership in science and technology,” said U.S. Secretary of Energy Chris Wright. “From achieving fusion ignition to advancing critical materials and high-performance computing, these scientists and engineers are tackling some of our nation’s most complex challenges and translating groundbreaking research into real-world capabilities. Their work strengthens our nation’s energy security, drives American innovation, and ensures the United States remains at the forefront of scientific discovery.”

One of Williams’s most widely recognized achievements is the creation of the Roofline Performance Model, a visual framework that helps developers easily identify hardware bottlenecks and optimize their code. The concept was born out of the challenges he faced when writing his doctoral thesis on automatic performance tuning. Struggling to explain why a numerical method worked well on one machine but poorly on another, he created a graphical representation for his advisor. What started as a simple visual aid to avoid a wordy explanation quickly revealed its deep potential.

“Sam’s development of the Roofline model has been absolutely transformative. Before Roofline, developers often struggled to understand exactly why their applications weren’t performing as they expected,” said Jonathan Carter, Associate Laboratory Director for Computing Sciences at Berkeley Lab. “By providing a clear, visual map of hardware bottlenecks, his work has demystified complex constraints and empowers researchers to push the boundaries of what our supercomputers can achieve.”

The Roofline Model is now an industry-standard framework adopted globally by academia, industry, and DOE user facilities. Major technology vendors, including Intel and NVIDIA, have integrated these methodologies into their flagship performance analysis tools, giving millions of software developers powerful optimization capabilities. Originally rooted in traditional HPC, the model’s versatility has driven its expansion into machine learning, computer architecture design, computer-aided engineering, computer science education, and even options pricing in the financial sector.

As supercomputing architectures evolved, Williams continually expanded the Roofline model’s capabilities. He updated the framework to capture the intricate memory and processing layers of modern GPUs. While originally designed to analyze individual applications, the model was recently expanded to evaluate the massive, diverse scientific workload of an entire supercomputer, demonstrating this approach on NERSC’s Perlmutter system. He also extended this methodology to help researchers pinpoint bottlenecks across complex, multi-step scientific workflows.

Beyond performance modeling, Williams has led multiple efforts to increase the efficiency and productivity of supercomputing code. Notably, under the DOE X-Stack program, he collaborated with Mary Hall at the University of Utah on the development of stencil-based auto-tuning and communication-avoiding techniques. Under the subsequent DOE Exascale Computing Project (ECP) PROTEAS-TUNE, he conceptualized and built the progenitor brick-based stencil computations. Researchers at the University of Utah ultimately refined these into the Bricks library — a software tool that reorganizes complex data so it can be processed much faster, enabling major scientific applications to run seamlessly and efficiently across a variety of CPUs and GPUs. 

Williams also leads major efforts in novel architectural hardware and hardware-software co-design. He directed Berkeley Lab’s contributions to the ECP Hardware Evaluation project and served as a member of its PathForward project. Through numerous collaborations with NERSC, he evaluates next-generation experimental hardware, including field-programmable gate arrays (FPGAs), disaggregated memory architectures, and advanced microchips, to maximize performance while operating under realistic power limits on supercomputers. These collaborations have resulted in multiple Best Paper awards. 

“What makes Sam’s work so impactful is his continuous drive to look ahead. From providing entirely new ways to understand performance bottlenecks to his current research integrating machine learning with high-performance computing, Sam’s contributions continue to advance the field,” said Stefan Wild, Director of the Applied Mathematics and Computational Research Division. “Sam is an outstanding researcher, and we are immensely proud to celebrate his selection for this prestigious award.”

Reflecting the spirit of large-scale, collaborative science that E. O. Lawrence championed, Williams noted that his success is the result of years of dedicated teamwork.

“I am incredibly honored by this recognition, but this body of work is truly the result of a massive team effort,” Williams said. “From the early funding risks taken by lab and ASCR leadership to the continuous collaborations within Berkeley Lab’s Computing Sciences divisions, NERSC, the broader Department of Energy complex, and our partners in industry and academia, none of this would have been possible without that collaborative support.”

###

Lawrence Berkeley National Laboratory (Berkeley Lab) is committed to groundbreaking research focused on discovery science and solutions for abundant and reliable energy supplies. The lab’s expertise spans materials, chemistry, physics, biology, earth and environmental science, mathematics, and computing. Researchers from around the world rely on the lab’s world-class scientific facilities for their own pioneering research. Founded in 1931 on the belief that the biggest problems are best addressed by teams, Berkeley Lab and its scientists have been recognized with 17 Nobel Prizes. Berkeley Lab is a multiprogram national laboratory managed by the University of California for the U.S. Department of Energy’s Office of Science.

DOE’s Office of Science is the single largest supporter of basic research in the physical sciences in the United States, and is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science.