2014 Poster Sessions : An FPGA Programming System for DNA Sequence Alignment

Student Name : Albert Ng
Advisor : William Dally
Research Areas: Computer Systems
FPGAs have been shown to be extremely effective at accelerating popular dynamic programming DNA sequence alignment algorithms, such as Smith-Waterman and HMMer. However, the difficulty in programmability poses too large of a barrier for typical bioinformaticists to use FPGAs when developing their tools. In this work, we develop a domain-specific high-level language for dynamic programming sequence alignment. We then develop a compiler tool-chain that allows the bioinformaticist to automatically generate FPGA accelerated implementations from the high-level language description.

Albert Ng is a PhD candidate in Electrical Engineering at Stanford University, pursuing research on the boundary of Computer Architecture and Bioinformatics under Dr. William Dally (EE/CS) and Dr. Gill Bejerano (CS/DevBio). He received his MS in EE at Stanford in 2013 and received his BS in EE from Caltech in 2011. He is the recipient of the Microsoft Research PhD Fellowship, the National Science Foundation Graduate Research Fellowship, and the Stanford Electrical Engineering Fellowship. Albert can be contacted at albertng@stanford.edu.