Generating Parallel Transforms Using Spiral

September 20, 2006
Franz Franchetti and Markus Püsche | Carnegie Mellon University, ECE Department

Spiral (www.spiral.net) is a program generation system for linear transforms such as the discrete Fourier transform, discrete cosine transforms, filters, and others. For a user-selected transform, Spiral autonomously generates different algorithms, represented in a declarative form as mathematical formulas, and their implementations to find the best match to the given target platform. Besides the search, Spiral performs deterministic optimizations on the formula level effectively restructuring the code in ways unpractical at the code level.

In this talk we give a short overview on Spiral and explain then how Spiral generates efficient programs for parallel platforms including vector architectures, shared and distributed memory platforms, and GPUs. As all optimizations in Spiral, parallelization

Speaker Details

Franz Franchetti is a Systems Scientist (special faculty) with the Deptartment of Electrical and Computer Engineering at Carnegie Mellon University. He received the Dipl.-Ing. (M.Sc.) degree and the Ph.D. degree in Technical Mathematics from the Vienna University of Technology in 2000 and 2003, respectively. He was a postdoctoral research associate with the Institute for Analysis and Scientific Computing during 2003. In 2004-2005 he was a postdoctoral research associate with the Department of Electrical and Computer Engineering at Carnegie Mellon University and a recipient of the Schroedinger fellowship awarded by the Austrian Science Fund. His research focuses on the development of high performance DSP algorithms. He is member of the SPIRAL team (www.spiral.net). More information is available at http://www.ece.cmu.edu/~franzf

Markus Püschel is an Associate Research Professor of Electrical and Computer Engineering at Carnegie Mellon University. He received his Diploma (M.Sc.) in Mathematics and his Doctorate (Ph.D.) in Computer Science, in 1995 and 1998, respectively, both from the University of Karlsruhe, Germany. From 1998-1999 he was a Postdoctoral Researcher in Mathematics and Computer Science at Drexel University. Since 2000 he has been with Carnegie Mellon University. He is an Associate Editor for the IEEE Transactions on Signal Processing, and was an Associate Editor for the IEEE Signal Processing Letters, a Guest Editor of the Journal of Symbolic Computation, and the Proceedings of the IEEE. He is a leader of the Spiral project (www.spiral.net) on automatic software/hardware generation of DSP algorithms. His research interests include scientific computing, compilers, applied mathematics and algebra, and signal processing theory/software/hardware. More information is available at www.ece.cmu.edu/~pueschel.