Abstract

This paper deals with scheduling, mapping and partitioning techniques for uniform loop nests. It is shown how the different techniques of scheduling, of mapping and of partitioning are linked and how code generation can be derived according to these methods. Our approach is based upon extensions of systolic array design methodologies.

Citations

405	Supercompilers for Parallel and Vector Computers - Zima, Chapman - 1990
300	Compiling Fortran D for MIMD distributed-memory machines - Hiranandani, Kennedy, et al. - 1992
192	Some efficient solutions to the affine scheduling problem, part 2 : multidimensional time - Feautrier - 1992
188	Dataflow analysis of array and scalar references - Feautrier - 1991
132	VLSI Array processors - Kung - 1988
69	Automatic synthesis of systolic arrays from uniform recurrent equations - Quinton - 1984
39	Minimim Distance: A method for partitioning recurrences for multiprocessors - Peir, Cytron - 1989
35	Systolic Algorithms and Architectures - Quinton, Robert - 1991
24	Regular partitioning for synthesizing fixed-size systolic arrays - Darte - 1991
23	An introduction to a formal theory of dependence analysis - Banerjee - 1988
17	A Graph--Theoretic Approach to the Alignment Problem - Darte, Robert - 1993
16	Loop nest scheduling and transformations - Darte, Risset, et al. - 1993
11	A ne-by-statement scheduling of uniform loop nests over parametric domains - Darte, Robert - 1992
10	Partitioning and mapping algorithms into fixed-size systolic arrays - Moldovan, Fortes - 1986
6	Languages and Compilers for Parallel Computing - Gelernter, Nicolau, et al. - 1990
2	A design methodology for fixed-size systolic arrays - B, Depreterre, et al. - 1990