Polaris: Improving the effectiveness of parallelizing compilers Conference Paper uri icon

abstract

  • © Springer-Verlag Berlin Heidelberg 1995. It is the goal of the Polaris project to develop a new parallelizing compiler that will overcome limitations of current compilers. While current parallelizing compilers may succeed on small kernels, they often fail to extract any meaningful parallelism from large applications. After a study of application, codes, it was concluded that by adding a few new techniques to current compilers, automatic parallelization becomes possible. The techniques needed are interprocedural analysis, scalar and array privatization, symbolic dependence analysis, and advanced induction and reduction recognition and elimination, along with run-time techniques to allow data dependent behavior.

name of conference

  • Languages and Compilers for Parallel Computing, 7th International Workshop, LCPC'94, Ithaca, NY, USA, August 8-10, 1994, Proceedings

published proceedings

  • Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

author list (cited authors)

  • Blume, W., Eigenmann, R., Faigin, K., Grout, J., Hoeflinger, J., Padua, D., ... Weatherford, S

citation count

  • 36

complete list of authors

  • Blume, William||Eigenmann, Rudolf||Faigin, Keith||Grout, John||Hoeflinger, Jay||Padua, David||Petersen, Paul||Pottenger, William||Rauchwerger, Lawrence||Tu, Peng||Weatherford, Stephen

editor list (cited editors)

  • Pingali, K., Banerjee, U., Gelernter, D., Nicolau, A., & Padua, D. A.

publication date

  • January 1995