Today's Roadrunner supercomputer is one trillion times faster than the MANIAC computer that was built at Los Alamos in 1952. The next generation of supercomputers, due around 2020, will provide another 1000-fold increase in performance. Los Alamos is currently investing in co-design projects that will both inform the design of these systems and ensure that our codes run productively and efficiently on them. Follow the links below for more information.
ExMatEx is one of three DoE ASCR-funded exascale co-design centers. The center focusses on extreme-scale simulation of materials properties in extreme environments. The project emphasizes multi-scale models, computer science, and validation in one co-design process designed to move materials simulation into the exascale era.
Visit the ExMatEx site.
CoCoMANS is an institutionally funded LDRD project at Los Alamos that is forging a qualitatively new predictive-science capability that exploits evolving high-performance computer architectures for multiple application areas—including materials, plasmas, and climate—by simultaneously evolving the four corners of science, methods, software, and hardware in an integrated computational co-design process.
Visit the CoCoMANS site.
The Los Alamos IS&T Co-Design Summer School was inaguarated in 2011 to recruit outstanding students to the laboratory for an exciting summer project. Students from diverse technical backgrounds including nuclear engineering, applied mathematics, and computer science, form teams that work together to solve a focussed co-design problem of interest to the laboratory. Their code and reports are published as open source.
Read more about the Summer School.
Optimization Principles is an institutionally funded LDRD project at Los Alamos that proposes to develop a framework for hardware-software co-design as a formally posed optimization problem. The project views co-design as search and selection from a vast space of hardware and software designs that map to performance metrics. The objective function to be optimized has as main components run time (or computational rate), problem size, simulated time duration, energy use, and hardware cost.
Visit the Optimization Principles site.
All of LANL's co-design projects have a significant open source component. Proxy applications—from micro-benchmarks to compact applicatons—will be released as open source code along with detailed documentation. These releases will enable us to more effectively collaborate with our industry and academic partners.
View the Los Alamos open source repository at GitHub.