Vaidy Sunderam: Resource and Application Adaptivity in Message Passing Systems
Clusters and MPP’s are traditional platforms for message passing applications, but there is growing interest in more dynamic metacomputing environments. The latter are characterized by dynamicity in availability and available capacity — of both nodes and interconnects. This talk will discuss fundamental challenges in executing message passing programs in such environments, and analyze the issue of adaptivity from the resource and application points of view. Pragmatic solutions to some of these challenges will then be described, along with new approaches to dealing with the aggregation of multidomain computing platforms for distributed memory concurrent computing.
About the speaker.
Vaidy Sunderam is a professor of Computer Science at Emory University. His research interests are in parallel and distributed processing systems, high-performance message passing software, and infrastructures for collaborative computing. His prior and current research efforts have focused on system architectures and implementations for network protocol middleware, collaboration tools, and heterogeneous metacomputing, including the PVM system and several other frameworks such as IceT, H2O, and Harness. Professor Sunderam teaches computer science at the beginning, advanced, and graduate levels, and advises graduate theses in the area of computer systems.