LOG IN TO MyLSU
Home
lecture image Special Guest Lectures
The development of a mixed mode programming paradigm
Meng-Shiou Wu, Postdoctoral Research Fellow,Scalable Computing Laboratory, Ames, Iowa
Applicant for CCT's CyD IT Analyst Position
Johnston Hall 338
July 09, 2007 - 10:00 am
Abstract:
In this presentation, Meng-Shiou Wu will talk about his involvement in several HPC projects since he began his work at Iowa State University in 2000. These include: the development of a mixed mode programming paradigm; the design of an automatically tuned collective communication system for SMP clusters; the construction of a performance evaluation system for GAMESS (General Atomic and Molecular Electronic Structure System), and designing mechanisms to enable GAMESS to interoperate with NWChem and MPQC under the Common Component Architecture (CCA).
div
Speaker's Bio:
Meng-Shiou Wu is currently a Postdoctoral Research Fellow at the Scalable Computing Laboratory, Ames Laboratory in Iowa. He received his Ph.D. in Computer Engineering from Iowa State University, his M.S. in Computer Science from the University of Denver, and a B.S. in Information Science from Tunghai University. Wu's research interests are in high performance computing. Currently this includes system-level optimizations of software, performance evaluation and prediction, and communication libraries. His Ph.D research is specialized in the performance analysis and optimizations of collective communications on SMP clusters. He developed interest in automatic tuning systems when working on tunable collective communications. Wu's research as a postdoctoral fellow since November 2005 has been focused on enhancing capability of computational chemistry software package GAMESS (General Atomic and Molecular Electronic Structure System). This includes integrating performance tools into GAMESS and constructing CCA (Common Component Architecture) components for GAMESS computations.