Abstract

The demand for high computing power has led to rapid deployment of large-scale parallel, cluster, and grid-based computing systems. A particular interest is that these computing systems enable parallelism to be exploited at all levels (thread, instruction, task, and program). However, investigating performance issues in such systems has proven to be a challenging task that requires the innovative performance analysis tools and techniques to keep up with the rapid evolution and ever increasing complexity of such systems. This talk highlights challenges, discusses techniques, and reports state-of-the-art and in-progress research of performance evaluation and measurement tools for assessing these computing systems.

Bio

Hong Ong is a Sr. HPC Evaluation Researcher at Oak Ridge National Laboratory (ORNL). Hong earned his Ph.D. from the University of Portsmouth, UK. Prior to joining ORNL, he worked on a number of grid-related projects, including the UK e-Science OGSA Testbed project and work evaluating security and firewall issues. His research interests include performance modeling and evaluation and distributed middleware for large-scale systems.