Supercomputing

Helping to advance and provide Rare Access to World-Class Computing Resources

The Oak Ridge Institute for Science Education (ORISE) recognizes that computational capabilities play an integral role in everything from stockpile stewardship and defense intelligence, to climate prediction and manufacturing. It is important to continue to advance high-performance computing, so that scientists can continue to perform mass computing activities at especially high speeds. 

ORISE supports the U.S. Department of Energy (DOE) and other federal agencies that are aiming to provide meaningful opportunities for faculty, students and other non-government officials to engage in research at world-class computing facilities.

ORISE assists NNSA in evaluating the World’s Fastest Machine

ORISE’s role: The 1.105-petaflop IBM supercomputer, located at Los Alamos National Laboratory (LANL) and nicknamed Roadrunner, has had the distinction of being rated the fastest supercomputer in the world—twice. While the system was still under development, the National Nuclear Security Administration (NNSA) wanted assurances that the project was on track and would be able to meet the scientific goals outlined in the original plan submitted by LANL. ORISE’s scientific peer review staff provided an objective program review to evaluate the progress of the supercomputer’s development.

Results: ORISE helped organize and facilitate a panel review where independent reviewers heard from vendors and LANL program managers about progress being made on the computer’s design, implementation and operation. As part of the Roadrunner’s operation goals, the supercomputer will aid in the assessment of aging of nuclear weapons through simulating and modeling programs, a method that will take the place of underground testing.

INCITE—A DOE Advanced Scientific Computing Program

ORISE’s role: ORISE coordinated a peer review of 80 proposals for DOE’s Innovation and Novel Computational Impact on Theory and Experiment (INCITE), which allocates processor-hours on some of the world’s most powerful supercomputers at DOE national laboratories. Proposals were evaluated for scientific merit, suitability of algorithms to perform well on a system’s architecture, and computational readiness of a code to make effective use of the system by using a majority of its processors in a single run.

Results: Helping to advance research in physics, chemical sciences, climate research, life sciences, material sciences, fusion, and nuclear engineering, the 66 project teams were awarded a total of 900 million hours, data storage resources, and the opportunity to work with computing experts.