Solving the constant-degree parallelism alignment problem
We describe an exact algorithm for finding a computation mapping and data distributions that minimize, for a given degree of parallelism, the number of remote data accesses in a distributed memory parallel computer (DMPC). This problem is shown to be NP-hard.
Unable to display preview. Download preview PDF.
- 2.J. M. Anderson and M. S. Lam. Global optimizations for parallelism and locality on scalable parallel machines. In Proc. PLDI '93, pages 112–125, 1993.Google Scholar
- 3.D. Bau, I. Kodukula, V. Kotlyar, K. Pingali, and P. Stodghill. Solving alignment using elementary linear algebra. In Proc. LCPC '94, LNCS 892, pages 46–60, Springer, 1994.Google Scholar
- 4.C. G. Diderich and M. Gengler. Solving the constant-degree parallelism alignment problem. Research Rep. DI-96/195, Swiss Fed. Inst. of Tech.-Lausanne, Switzerland, 1996.Google Scholar
- 5.Michèle Dion and Yves Robert. Mapping affine loop nests: New results. In Proc. HPCN '95, LNCS 919, pages 184–189, Springer, 1995.Google Scholar
- 6.C.-H. Huang and P. Sadayappan. Communication-free hyperplane partitioning of nested loops. In Proc. LCPC '91, LNCS 589, pages 186–200, Springer, 1991.Google Scholar