Guy Blelloch, Professor of Computer Science and Associate Dean for Undergraduate Education at School of Computer Science will be speaking at Carnegie Mellon University in Qatar on “Parallel Algorithms Come of Age”. 

To confirm your attendance, please RSVP.

About the talk: 

With the advent of multicore processors there are almost no computers or devices, beyond perhaps a toaster, that still only have a single processing core.  Furthermore it is easy and reasonably cheap to get machines with many cores.  And, these machines support algorithms that can run dozens or hundreds of times faster than their sequential counterparts.  However, most universities still only, or certainly mostly, teach sequential algorithms as part of their core undergraduate curriculum.  I’ll discuss some of the problems in teaching parallel algorithms, but why now is the right time to convert to teaching parallel algorithms and programming.  In particular the languages and tools for supporting parallelism have reached an adequate level of maturity.  Furthermore the understanding of parallel algorithms from a research point of view, including methods for analyzing asymptotic costs, is quite far along.  And, perhaps most importantly, parallel algorithms are now very much faster than sequential algorithms, and the the theory matches the practice. 

Biography: 

Guy E. Blelloch received his B.S. and B.A. from Swarthmore College in 1983, and his M.S. and PhD from MIT in 1986, and 1988, respectively. Since then he has been on the faculty at Carnegie Mellon University, where he is now an Associate Professor and Associate Dean for Undergraduate Education.  During the academic year 1997-1998 he was a visiting professor at U.C. Berkeley. He held the Finmeccanica Faculty Chair from 1991–1995 and received an NSF Young Investigator Award in 1992. He has been program chair for the ACM Symposium on Parallel Algorithms and Architectures, program co-chair for the IEEE International Parallel Processing Symposium, is on the editorial board of JACM, and has served on a dozen or so program committees. His research interests are in parallel programming languages and parallel algorithms, and in the interaction of the two. He has developed the NESL programming language under an ARPA contract and an NSF NYI award. His work on parallel algorithms includes work on sorting, computational geometry, and several pointer-based algorithms, including algorithms for list-ranking, set-operations, and graph connectivity.