Adaptive strategy for one-sided communication in MPICH2

Zhao, Xin

Permalink

https://hdl.handle.net/2142/31986 Copy

Description

Title

Adaptive strategy for one-sided communication in MPICH2

Author(s)

Zhao, Xin

Issue Date

2012-06-27T21:23:13Z

Director of Research (if dissertation) or Advisor (if thesis)

Gropp, William D.

Department of Study

Computer Science Computer Science

Discipline

Computer Science Computer Science

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

M.S.

Degree Level

Thesis

Keyword(s)

• One-sided Communication
• MPI implementation
• Synchronization
• Adaptive Strategy
• Message Passing Interface (MPI)

Abstract

The one-sided communication model (or remote memory access) supported by MPI-2 is more convenient to use than the regular two-sided send/receive communication model, because it allows the sender to specify all data transfer parameters and avoids the receiver to be explicitly involved in data receiving. One-sided communication model also has potential to provide higher performance. The MPI-2 standard provides flexibility about when RMA operations can be issued and completed, which makes the MPI implementation possible to be optimized internally. The current MPICH2 implementation uses a lazy approach to issue one-sided operations by queuing the operations and issuing them during the later synchronization phase. This has certain benefits with respect to short operations in terms of reduced network operations. For large data transfers, issuing operations in an eager fashion could be more beneficial as well as provide more scope for overlapping communication and computation. In this thesis we describe the design and implementation of an adaptive approach for all three synchronization mechanisms defined in MPI: fence, post-start-complete-wait, and lock-unlock. We evaluate our implementation with both micro benchmarks and the Graph500 benchmark to demonstrate the performance impact of our approach. The performance results show that our hybrid approach performs as good as the lazy approach for small data transfers and achieves performance similar to the eager approach for large data transfers, and its overlapping percentage is good.

Graduation Semester

2012-05

Permalink

http://hdl.handle.net/2142/31986

Copyright and License Information

Copyright 2012 Xin Zhao

Owning Collections