Boston University | Center for Computational Science
HomeNews and EventsResearchEducationPeopleSeminarsFacilitiesContact Us

"BlueGene/L: A new kind of Supercomputer"

David Klepacki, Ph.D. - IBM
April 8, 2003
 
This talk addresses IBM's future directions in High Performance Computing Technology. Specifically, it focuses on cellular architectures and in particular the BlueGene/L Supercomputer. BlueGene/L consists of 65,536 compute nodes with a target peak processing power of 360 teraflops (trillion floating-point operations per second) in the 2004-2005 timeframe. Its price/performance and power consumption/performance targets are unobtainable with conventional system architectures. Furthermore, these systems are designed to accommodate more general purpose technical computing and will use standard programming models, such as MPI message passing, along with standard high-level programming languages such as FORTRAN and C/C++.
 

Many Scientific disciplines are expected to benefit from this computing architecture, including fluid dynamics, chemistry, materials science, engineering, and the life sciences. A wide variety of applications are currently being studied via collaborations with Lawrence Livermore National Labs (LLNL), San Diego Supercomputing Center (SDSC) and the California Institute of Technology's Center for Advanced Computing Research (CACR)
 
Current analysis of many codes have shown that they will benefit in terms of scalability performance due to BlueGene/L's fast multiple networks and optimized message-passing support. The current status of this project (hardware, software and application analysis) will be presented.
 
David Klepacki is a senior staff member of IBM Research and has more than 20 years of experience in high performance computing. He has worked in the areas of high-performance processor design, numerically intensive computation, computational physics, parallel computing, application benchmarking and cluster computing. Currently, he directs and manages the Advanced Computing Technology Center (ACTC) at IBM Research in Yorktown Heights, New York. This center has a mission to study performance modeling as well as to provide application optimization services to the IBM HPC user community. David earned his Ph.D. in Theoretical Nuclear Physics from Purdue University at West Lafayette, Indiana. He also has a Masters degree in Electrical and Computer Engineering from Syracuse University in New York.
 

copyright © 2006, Center for Computational Science | Boston University , MA, 02215