A consistent generation of pipeline parallelism and distribution of operations and data among processors

Authors:
E. V. Adutskevich;N. A. Likhoded
Affiliations:
Institute of Mathematics, National Academy of Sciences of Belarus, Minsk, Belarus 220072;Institute of Mathematics, National Academy of Sciences of Belarus, Minsk, Belarus 220072
Venue:
Programming and Computing Software
Year:
2006

Citing 10
Cited 0

Some efficient solutions to the affine scheduling problem: I. One-dimensional time

International Journal of Parallel Programming
Mapping uniform loop nests onto distributed memory architectures

Parallel Computing
Affine-by-statement scheduling of uniform and affine loop nests over parametric domains

Journal of Parallel and Distributed Computing
Mapping affine loop nests

Parallel Computing
Maximizing parallelism and minimizing synchronization with affine partitions

Parallel Computing - Special issues on languages and compilers for parallel computers
Automatic Generation of Modular Time-Space Mappings and Data Alignments

Journal of VLSI Signal Processing Systems - Special issue on application specific systems, architectures and processors
An affine partitioning algorithm to maximize parallelism and minimize communication

ICS '99 Proceedings of the 13th international conference on Supercomputing
Blocking and array contraction across arbitrarily nested loops using affine partitioning

PPoPP '01 Proceedings of the eighth ACM SIGPLAN symposium on Principles and practices of parallel programming
A Method for Parallelizing Algorithms by Vector Scheduling Functions

Programming and Computing Software
Distribution of Operations and Data Arrays over Processors

Programming and Computing Software

Quantified Score

Hi-index	0.00

Visualization

Abstract

The problem of mapping affine loop nests onto parallel computers with distributed memory is considered. A technique for algorithm scheduling and distributing operations and data over processors is proposed. This technique makes it possible to generate pipeline parallelism and minimize the amount of data exchanges between the processors. The method is adapted for automation and explicitly allows for dependence on outer variables of loops.