Improving MPI communication via data type fission

Authors:
Ben Perry;Martin Swany
Affiliations:
University of Delaware, Newark, DE;University of Delaware, Newark, DE
Venue:
Proceedings of the 19th ACM International Symposium on High Performance Distributed Computing
Year:
2010

Citing 12
Cited 0

Compiler transformations for high-performance computing

ACM Computing Surveys (CSUR)
Improving Cache Locality by a Combination of Loop and Data Transformations

IEEE Transactions on Computers - Special issue on cache memory and related problems
Improving cache performance in dynamic applications through data and computation reorganization at run time

Proceedings of the ACM SIGPLAN 1999 conference on Programming language design and implementation
MPI-The Complete Reference, Volume 1: The MPI Core

MPI-The Complete Reference, Volume 1: The MPI Core
MPI-2: Extending the Message-Passing Interface

Euro-Par '96 Proceedings of the Second International Euro-Par Conference on Parallel Processing - Volume I
An MPI Implementation of the BLACS

HIPC '96 Proceedings of the Third International Conference on High-Performance Computing (HiPC '96)
LLVM: A Compilation Framework for Lifelong Program Analysis & Transformation

Proceedings of the international symposium on Code generation and optimization: feedback-directed and runtime optimization
Array regrouping and structure splitting using whole-program reference affinity

Proceedings of the ACM SIGPLAN 2004 conference on Programming language design and implementation
Practical Structure Layout Optimization and Advice

Proceedings of the International Symposium on Code Generation and Optimization
Open MPI community meeting

Proceedings of the 2006 ACM/IEEE conference on Supercomputing
Shape analysis for composite data structures

CAV'07 Proceedings of the 19th international conference on Computer aided verification
Automatic Realignment of Data Structures to Improve MPI Performance

ICN '10 Proceedings of the 2010 Ninth International Conference on Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

Message Passing Interface (MPI) messages are centered around transmitting instances of MPI data types. The data types represented in MPI terms are usually modeled after data types native to the application. If a user does not want to transmit a field from the native data type, the user will sometimes align the MPI data type such that there is a gap in the displacement where the omitted field would be. With the resulting MPI data type now being non-contiguous, cycles are spent making the data contiguous for transmission and then expanding the data back out on the receiving side. We show that by performing data type fission (the process of segregating the transmitted fields from the non-transmitted fields) and aligning the MPI data type accordingly, we can completely eliminate the need to copy data during the packing and unpacking process, which can significantly improve the performance of communication-heavy Single Program Multiple Data applications.