DEPARTMENT Home About Us Contact Us People Research Careers & Employment INFORMATION MENU Admissions Information Courses Organizations Undergrad Program Information Graduate Program Information Scholarships/Awards Student Events Resources & Labs ACTION MENU Login To See Your Actions

Department Seminar Series

2009-2010 Seminar Series

High Performance Computing at the Oak Ridge National Laboratory

Speaker: Dr. Olaf O. Storaasli
Date: September 28th, 2009
Time: 3:30pm
Place: Thorvaldson 105

For more information on this, and other seminars in the Department Seminar Series, please see the Department Seminar Series page.

Abstract

In the '70's, the "coin of the realm" in high-performance computing was to achieve Millions of FLoating Point Operations/Sec (MFLOPS) to solve scientific applications. In 1989, the next level of performance (1,000 MFLOPS) was achieved by a NASA-ODU team (Storaasli-Nguyen-Agarwal) who received Cray's 1st GigaFLOP Award for the static analysis of a "gigantic" Space Shuttle Solid Rocket Booster model with 54K equations. In 1998, an ORNL team using a Cray T3E broke the TeraFLOP (million MFLOPS) barrier. In 2008, two DOE computers (ORNL Jaguar & LANL Roadrunner) exceeded 1 PetaFLOP (billion MFLOPS), a phenomenal billion times faster than our early NASA leading-edge computers.

Along with theory and experiment, the DOE views supercomputers as the third leg of scientific discovery. Further, as experimentation is costly and slow, an increasing number of significant discoveries are made using supercomputers.

Dr Storaasli will describe the current top fleet of supercomputers (mostly DOE) and applications they are now solving. Based on his research in the Future Technologies Group, he will project, the next level of supercomputers (ExaFLOP): their architecture, software, tools and the performance they are expected to achieve to enable unparalleled scientific discoveries.

Biography

In 2005, after his 35-year career at NASA Langley, Dr. Storaasli joined CSM's Future Technologies Group to explore new algorithms and architectures to harness the power of Field Programmable Gate Arrays (FPGAs) for future scientific supercomputing (including rapid solution of large matrix equations). Olaf earned his Ph.D. in Engineering Mechanics from NCSU. (More info: http://ft.ornl.gov/~olaf)

Return to Department Seminar Series