Mary Hall: Compiling for high-performance architectures

Reading Assignments:

Evaluation of Predicated Array Data-Flow Analysis for Automatic Parallelization, S. Moon and M. W. Hall, In Proceedings of the ACM Symposium on Principles and Practice of Parallel Programming, May, 1999.

Mapping Irregular Computations to DIVA, a Data-Intensive Architecture, Mary Hall, Peter Kogge, Jeff Koller, Pedro Diniz, Jacqueline Chame, Jeff Draper, Jeff LaCoss, John Granacki, Jay Brockman, William Athas, Apoorv Srivastava, Jaewook Shin, Joonseok Park, Proceedings of SC99, Nov. 1999.

"A Compiler Approach to Fast Design Space Exploration for FPGA-based Systems," B. So, M. Hall and P. Diniz, Proceedings of the ACM Conference on Programming Language Design and Implementation '02, June, 2002.

Three questions: (pick one and answer it in a one-page essay that you will hand in to Prof. Itti the monday following the talk, at noon in class):

- What are the technology trends that have influenced computer architecture,
and what impact have these had on compiler technology?
- When compiling to an FPGA-based system, what defines an "optimal" solution
and why?
- What are the advantages and disadvantages of an empirical optimization
strategy?