An affine partitioning algorithm to maximize parallelism and minimize communication
The DOI matches its Crossref record; 16 of the 16 references it cites were found in a registry or an index.
Abstract
An affine partitioning framework unifies many useful program transforms such as unimodular transformations (interchange, reversal, skewing), loop fusion, fission, scaling, reindexing, and statement reordering. This paper presents an algorithm, based on this unified framework, that maximizes parallelism while minimizing communication in programs with arbitrary loop nestings and affine data accesses. Our algorithm can find the optimal affine partition that maximizes the degree of parallelism with the minimum degree of synchronizations. In addition, it uses a greedy algorithm to minimize communication between loops heuristically by aligning the computation partitions for different loops, trading off excess degrees of parallelism, and choosing pipelined parallelism over doall parallelism if it can significantly reduce the communication. The algorithm is optimal in maximizing the degrees of parallelism that require (1) no communication, (2) near-neighbor communication and a constant number ...
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Theory of Linear and Integer Programming
Letchford, A. N., & Schrijver, A. (2000) · The Journal of the Operational Research Society · doi:10.2307/253980 · Cited by 5,515
Theory of Linear and Integer Programming
Smith, D. K. (1987) · Journal of the Operational Research Society · doi:10.1057/jors.1987.94 · Cited by 2,708
Optimizing supercompilers for supercomputers
Wolfe, M. (2020) · Proceedings of the 34th ACM International Conference on Supercomputing · doi:10.1145/3392717.3400034 · Cited by 837
Advanced compiler optimizations for supercomputers
Padua, D. A., & Wolfe, M. J. (1986) · Communications of the ACM · doi:10.1145/7902.7904 · Cited by 690
Dependence graphs and compiler optimizations
Kuck, D. J., Kuhn, R. H., Padua, D. A., Leasure, B., & Wolfe, M. (1981) · Proceedings of the 8th ACM SIGPLAN-SIGACT symposium on Principles of programming languages - POPL '81 · doi:10.1145/567532.567555 · Cited by 651 · cited 2×
“5 Comparison with Other Algorithms The a ne partitioning framework uni es a large number of previously proposed loop transformations, including unimodular transformations (interchange, reversal, skewing), reindexing, distribution, fusion, scaling, and statement reordering[1, 5, 11, 13, 17, 25].”
BackgroundMethodA loop transformation theory and an algorithm to maximize parallelism
Wolf, M. E., & Lam, M. S. (1991) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/71.97902 · Cited by 643 · cited 2×
“Unimodular transformations (equivalent to arbitrary combinations of loop interchange, reversal and skewing) and blocking/tiling (equivalent to unroll-and-jam or stripmine-and-interchange) are e ective for parallelizing perfectly nested loops[4, 5, 23, 24, 25].”
BackgroundSelf-citationSome efficient solutions to the affine scheduling problem. I. One-dimensional time
Feautrier, P. (1992) · International Journal of Parallel Programming · doi:10.1007/bf01407835 · Cited by 555
Some efficient solutions to the affine scheduling problem. Part II. Multidimensional time
Feautrier, P. (1992) · International Journal of Parallel Programming · doi:10.1007/bf01379404 · Cited by 368
Data-centric multi-level blocking
Kodukula, I., Ahmed, N., & Pingali, K. (1997) · Proceedings of the ACM SIGPLAN 1997 conference on Programming language design and implementation · doi:10.1145/258915.258946 · Cited by 238 · cited 1×
“proposed the idea of data shackling whereby a compiler uses the owner-compute rule to distribute a program with arbitrary loop nests and sequences according to the user's data distribution directives[12].”
MethodMaximizing parallelism and minimizing synchronization with affine transforms
Lim, A. W., & Lam, M. S. (1997) · Proceedings of the 24th ACM SIGPLAN-SIGACT symposium on Principles of programming languages - POPL '97 · doi:10.1145/263699.263719 · Cited by 179
Self-citation
Cited by
A practical automatic polyhedral parallelizer and locality optimizer
Bondhugula, U., Hartono, A., Ramanujam, J., & Sadayappan, P. (2008) · Proceedings of the 29th ACM SIGPLAN Conference on Programming Language Design and Implementation · doi:10.1145/1375581.1375595 · Cited by 880
A practical automatic polyhedral parallelizer and locality optimizer
Bondhugula, U., Hartono, A., Ramanujam, J., & Sadayappan, P. (2008) · ACM SIGPLAN Notices · doi:10.1145/1379022.1375595 · Cited by 248
Blocking and array contraction across arbitrarily nested loops using affine partitioning
Lim, A. W., Liao, S.-W., & Lam, M. S. (2001) · Proceedings of the eighth ACM SIGPLAN symposium on Principles and practices of parallel programming · doi:10.1145/379539.379586 · Cited by 142
Self-citation
Metadata
- Authors
- Lim, A. W., Cheong, G. I., & Lam, M. S.
- Venue
- Proceedings of the 13th international conference on Supercomputing
- Publisher
- ACM
- Published
- 1999-05
- Pages
- 228-237
- OpenAlex
- W1964354616
- Cited by
- 181
- References
- 16
- FWCI
- 3.42
- Type
- conference paper
- Language
- en
- Open access
- closed
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.