Using Simulation to Examine the Effect of MPI Message Matching Costs on Application Performance

Levy, Scott; Ferreira, Kurt B.; Schonbein, Whit; Grant, Ryan E.; Dosanjh, Matthew G. F.

doi:10.1016/j.parco.2019.02.008

Title: Using Simulation to Examine the Effect of MPI Message Matching Costs on Application Performance

Full Record
Other Related Research

Abstract

Attaining high performance with MPI applications requires efficient message matching to minimize message processing overheads and the latency these overheads introduce into application communication. We use a validated simulation-based approach to examine the relationship between MPI message matching performance and application time-to-solution. Specifically, we examine how the performance of several important HPC workloads is affected by the time required for matching. Our analysis yields several important contributions: (i) the performance of current workloads is unlikely to be significantly affected by MPI matching unless match queue operations get much slower or match queues get much longer; (ii) match queue designs that provide sublinear performance as a function of queue length are unlikely to yield much benefit unless match queue lengths increase dramatically; and (iii) we provide guidance on how long the mean time per match attempt may be without significantly affecting application performance. The results and analysis in this paper provide valuable guidance on the design and development of MPI message match queues.

Authors:

Levy, Scott ^[1]; Ferreira, Kurt B. ^[1]; Schonbein, Whit ^[1]; Grant, Ryan E. ^[1]; Dosanjh, Matthew G. F. ^[1]

Sandia National Lab. (SNL-NM), Albuquerque, NM (United States). Center for Computing Research

Publication Date:: Wed Feb 27 00:00:00 EST 2019

Research Org.:: Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)

Sponsoring Org.:: USDOE National Nuclear Security Administration (NNSA)

OSTI Identifier:: 1502976

Alternate Identifier(s):: OSTI ID: 1636805

Report Number(s):: SAND-2019-3029J
Journal ID: ISSN 0167-8191; 673544

Grant/Contract Number:: NA0003525

Resource Type:: Journal Article: Accepted Manuscript

Journal Name:: Parallel Computing

Additional Journal Information:: Journal Volume: 84; Journal ID: ISSN 0167-8191

Publisher:: Elsevier

Country of Publication:: United States

Language:: English

Subject:: 97 MATHEMATICS AND COMPUTING

Citation Formats


                    Levy, Scott, Ferreira, Kurt B., Schonbein, Whit, Grant, Ryan E., and Dosanjh, Matthew G. F. Using Simulation to Examine the Effect of MPI Message Matching Costs on Application Performance.  United States: N. p., 2019. 
        Web.  doi:10.1016/j.parco.2019.02.008.

Copy to clipboard


                    Levy, Scott, Ferreira, Kurt B., Schonbein, Whit, Grant, Ryan E., & Dosanjh, Matthew G. F. Using Simulation to Examine the Effect of MPI Message Matching Costs on Application Performance.  United States.  https://doi.org/10.1016/j.parco.2019.02.008

Copy to clipboard


                    Levy, Scott, Ferreira, Kurt B., Schonbein, Whit, Grant, Ryan E., and Dosanjh, Matthew G. F. 2019.  
        "Using Simulation to Examine the Effect of MPI Message Matching Costs on Application Performance".  United States.  https://doi.org/10.1016/j.parco.2019.02.008.  https://www.osti.gov/servlets/purl/1502976.

Copy to clipboard


                    
@article{osti_1502976,

  title        = {Using Simulation to Examine the Effect of MPI Message Matching Costs on Application Performance},

  author       = {Levy, Scott and Ferreira, Kurt B. and Schonbein, Whit and Grant, Ryan E. and Dosanjh, Matthew G. F.},

  abstractNote = {Attaining high performance with MPI applications requires efficient message matching to minimize message processing overheads and the latency these overheads introduce into application communication. We use a validated simulation-based approach to examine the relationship between MPI message matching performance and application time-to-solution. Specifically, we examine how the performance of several important HPC workloads is affected by the time required for matching. Our analysis yields several important contributions: (i) the performance of current workloads is unlikely to be significantly affected by MPI matching unless match queue operations get much slower or match queues get much longer; (ii) match queue designs that provide sublinear performance as a function of queue length are unlikely to yield much benefit unless match queue lengths increase dramatically; and (iii) we provide guidance on how long the mean time per match attempt may be without significantly affecting application performance. The results and analysis in this paper provide valuable guidance on the design and development of MPI message match queues.},

  doi          = {10.1016/j.parco.2019.02.008},

  url          = {https://www.osti.gov/biblio/1502976},
  journal      = {Parallel Computing},

  issn         = {0167-8191},

  number       = ,

  volume       = 84,

  place        = {United States},

  year         = {Wed Feb 27 00:00:00 EST 2019},

  month        = {Wed Feb 27 00:00:00 EST 2019}

}

Copy to clipboard

Journal Article:

Free Publicly Available Full Text

Accepted Manuscript (Publisher)

Accepted Manuscript (DOE)

Publisher's Version of Record

https://doi.org/10.1016/j.parco.2019.02.008

Other availability

Search WorldCat to find libraries that may hold this journal

Citation Metrics:

Cited by: 1 work

Citation information provided by
Web of Science

Save / Share:

Export Metadata

Save to My Library

Similar records in OSTI.GOV collections:

Evaluating MPI resource usage summary statistics

Journal Article Ferreira, Kurt; Levy, Scott - Parallel Computing

The Message Passing Interface (MPI) remains the dominant programming model for scientific applications running on today’s high-performance computing (HPC) systems. This dominance stems from MPI’s powerful semantics for inter-process communication that has enabled scientists to write applications for simulating important physical phenomena. MPI does not, however, specify how messages and synchronization should be carried out. Those details are typically dependent on low-level architecture details and the message characteristics of the application. Therefore, analyzing an application’s MPI resource usage is critical to tuning MPI’s performance on a particular platform. The result of this analysis is typically a discussion of the meanmore »« less
https://doi.org/10.1016/j.parco.2021.102825

Full Text Available
Characterizing MPI matching via trace-based simulation

Journal Article Ferreira, Kurt; Levy, Scott; Pedretti, Kevin; ... - Parallel Computing

With the increased scale expected on future leadership-class systems, detailed information about the resource usage and performance of MPI message matching provides important insights into how to maintain application performance on next-generation systems. However, obtaining MPI message matching performance data is often not possible without significant effort. A common approach is to instrument an MPI implementation to collect relevant statistics. While this approach can provide important data, collecting matching data at runtime perturbs the application's execution, including its matching performance, and is highly dependent on the MPI library's matchlist implementation. In this paper, we introduce a trace-based simulation approach tomore »« less
https://doi.org/10.1145/3127024.3127040

Full Text Available
A new deadlock resolution protocol and message matching algorithm for the extreme-scale simulator

Journal Article Engelmann, Christian; Naughton, III, Thomas - Concurrency and Computation. Practice and Experience

Summary Investigating the performance of parallel applications at scale on future high‐performance computing (HPC) architectures and the performance impact of different HPC architecture choices is an important component of HPC hardware/software co‐design. The Extreme‐scale Simulator (xSim) is a simulation toolkit for investigating the performance of parallel applications at scale. xSim scales to millions of simulated Message Passing Interface (MPI) processes. The xSim toolkit strives to limit simulation overheads in order to maintain performance and productivity criteria. This paper documents two improvements to xSim: (1) a new deadlock resolution protocol to reduce the parallel discrete event simulation overhead and (2) a new simulated MPImore »« less
Cited by 1
https://doi.org/10.1002/cpe.3805

Full Text Available
A dynamic, unified design for dedicated message matching engines for collective and point-to-point communications

Journal Article Ghazimirsaeed, S.; Grant, Ryan; Afsahi, Ahmad - Parallel Computing

The Message Passing Interface (MPI) libraries use message queues to guarantee correct message ordering between communicating processes. Message queues are in the critical path of MPI communications and thus, the performance of message queue operations can have significant impact on the performance of applications. Collective communications are widely used in MPI applications and they can have considerable impact on generating long message queues. In this paper, we propose a unified message matching mechanism that improves the message queue search time by distinguishing messages coming from point-to-point and collective communications and using a distinct message queue data structure for them. Formore »« less
Cited by 2
https://doi.org/10.1016/j.parco.2019.102547

Full Text Available
Production Application Performance Data Streaming for System Monitoring

Journal Article Izadpanah, Ramin; Allan, Benjamin; Dechev, Damian; ... - ACM Transactions on Modeling and Performance Evaluation of Computing Systems

In this article, we present an approach to streaming collection of application performance data. Practical application performance tuning and troubleshooting in production high-performance computing (HPC) environments requires an understanding of how applications interact with the platform, including (but not limited to) parallel programming libraries such as Message Passing Interface (MPI). Several profiling and tracing tools exist that collect heavy runtime data traces either in memory (released only at application exit) or on a file system (imposing an I/O load that may interfere with the performance being measured). Although these approaches are beneficial in development stages and post-run analysis, a systemwidemore »« less
https://doi.org/10.1145/3319498

Full Text Available

Similar Records