Best Paper
Thursday, April 1st, 2010Computer Science doctoral student Bryan Franklin and Professor Steven Seidel received the Best Paper Award for their paper, “A Parallel Longest Common Subsequence Algorithm in UPC”, at the High Performance Computing Symposium, April 12-14, in Orlando, Florida. Franklin presented the paper at the conference. The paper describes the design, implementation, and performance of a parallel algorithm for the longest common subsequence problem, an important problem in bioinformatics. This is the second consecutive year that Michigan Tech authors have won the best paper award at the HPC Symposium.
The LCS algorithm was expressed in the new programming language, UPC. UPC is based on C and expresses parallel computation in a partitioned global address space. Languages such as UPC are being developed to make programming the coming generation of peta-scale supercomputers easier and more reliable. The UPC implementation developed in this work is simpler than a similar implementation using MPI, which is currently the most widely used way to express parallel algorithms.