Efficient MPI support for advanced hybrid programming models

Authors:
Torsten Hoefler;Greg Bronevetsky;Brian Barrett;Bronis R. De Supinski;Andrew Lumsdaine
Affiliations:
University of Illinois at Urbana-Champaign, Urbana, IL;Lawrence Livermore National Laboratory, Center for Applied Scientific Computing, Livermore, CA;Sandia National Laboratories, Albuquerque, NM;Lawrence Livermore National Laboratory, Center for Applied Scientific Computing, Livermore, CA;Indiana University, Open Systems Lab, Bloomington, IN
Venue:
EuroMPI'10 Proceedings of the 17th European MPI users' group meeting conference on Recent advances in the message passing interface
Year:
2010

Citing 6
Cited 1

Scalability of hybrid programming for a CFD code on the earth simulator

Parallel Computing
Design and implementation of a high-performance MPI for C# and the common language infrastructure

Proceedings of the 13th ACM SIGPLAN Symposium on Principles and practice of parallel programming
Toward Efficient Support for Multithreaded MPI Communication

Proceedings of the 15th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface
Issues in developing a thread-safe MPI implementation

EuroPVM/MPI'06 Proceedings of the 13th European PVM/MPI User's Group conference on Recent advances in parallel virtual machine and message passing interface
Test suite for evaluating performance of MPI implementations that support MPI_THREAD_MULTIPLE

PVM/MPI'07 Proceedings of the 14th European conference on Recent Advances in Parallel Virtual Machine and Message Passing Interface
Netgauge: a network performance measurement framework

HPCC'07 Proceedings of the Third international conference on High Performance Computing and Communications

Writing parallel libraries with MPI - common practice, issues, and extensions

EuroMPI'11 Proceedings of the 18th European MPI Users' Group conference on Recent advances in the message passing interface

Quantified Score

Hi-index	0.00

Visualization

Abstract

The number of multithreaded Message Passing Interface (MPI) implementations and applications is increasing rapidly. We discuss how multithreaded applications can receive messages of unknown size. As is well known, combining MPI_Probe/MPI_Recv is not threadsafe, but many assume that trivial workarounds exist. We discuss those workarounds and show how they fail in practice by either limiting the available parallelism unnecessarily, consuming resources in a nonscalable way, or promoting global deadlocks. In this light, we propose two fundamentally different efficient approaches to enable thread-safe messaging in MPI-2.2: fine-grained locking and matching outside of MPI. Our approaches provide thread-safe probe and receive functionality, but both have deficiencies, including performance limitations and programming complexity, that could be avoided if MPI would offer a thread-safe (stateless) interface to MPI Probe. We propose such an extension for the upcoming MPI-3 standard, provide a reference implementation, and demonstrate significant performance benefits.