November 22, 2017: Using predictive atomistic calculations and high-performance supercomputers at NERSC, University of Michigan researchers found that incorporating the element boron into the widely used InGaN (indium-gallium nitride) material can keep electrons from becoming too crowded in LEDs, making the material more efficient at producing light. Read More »
- Exascale and Beyond We're rethinking every aspect of scientific computing—hardware, software, algorithms, efficiency and networking—to address limits on processor speed, memory and energy consumption. The goal: more science per Watt.
- Data-Driven Science We develop tools and technologies and provide the critical networking and computer resources that help scientists turn enormous, widely shared, and complex data sets into discoveries.
- Scientific Complexity Engaging scientists in a holistic, team-based approach—from theory, observations and experiments to simulation and data analysis—is a grand challenge and goal of Computing Sciences at Berkeley Lab.
November 21, 2017: After nearly five years of collaboration between researchers in academia, industry and national research laboratories—including Berkeley Lab's Aydın Buluç—GraphBLAS, a collection of standardized building blocks for graph algorithms in the language of linear algebra, is publicly available. Read More »
November 10, 2017: In this Q&A with Prabhat, who leads the Data and Analytics Services Group at NERSC and has been instrumental in several projects exploring opportunities for deep learning in science, he talks about the history of deep learning and machine learning and the unique challenges of applying these data analytics tools to science. Read More »