Dynamic thread mapping based on machine learning for transactional memory applications

Authors:
M$#225;rcio Castro;Luís Fabrício Wanderley Góes;Luiz Gustavo Fernandes;Jean-Fran$#231;ois Méhaut
Affiliations:
INRIA - CEA - LIG Laboratory, Grenoble University, Montbonnot Saint Martin, France;Department of Computer Science, Pontifical Catholic University of Minas Gerais, Belo Horizonte, MG, Brazil;PPGCC - Pontifical Catholic University of Rio Grande do Sul, Porto Alegre, RS, Brazil;INRIA - CEA - LIG Laboratory, Grenoble University, Montbonnot Saint Martin, France
Venue:
Euro-Par'12 Proceedings of the 18th international conference on Parallel Processing
Year:
2012

Citing 12
Cited 1

Induction of Decision Trees

Machine Learning
Dynamic performance tuning of word-based software transactional memory

Proceedings of the 13th ACM SIGPLAN Symposium on Principles and practice of parallel programming
Mapping parallelism to multi-cores: a machine learning based approach

Proceedings of the 14th ACM SIGPLAN symposium on Principles and practice of parallel programming
Towards a holistic approach to auto-parallelization: integrating profile-driven parallelism detection and machine-learning based mapping

Proceedings of the 2009 ACM SIGPLAN conference on Programming language design and implementation
Process Mapping for MPI Collective Communications

Euro-Par '09 Proceedings of the 15th International Euro-Par Conference on Parallel Processing
hwloc: A Generic Framework for Managing Hardware Affinities in HPC Applications

PDP '10 Proceedings of the 2010 18th Euromicro Conference on Parallel, Distributed and Network-based Processing
Process variation aware thread mapping for chip multiprocessors

Proceedings of the Conference on Design, Automation and Test in Europe
Evaluating Thread Placement Based on Memory Access Patterns for Multi-core Processors

HPCC '10 Proceedings of the 2010 IEEE 12th International Conference on High Performance Computing and Communications
Eigenbench: A simple exploration tool for orthogonal TM characteristics

IISWC '10 Proceedings of the IEEE International Symposium on Workload Characterization (IISWC'10)
Analysis and Tracing of Applications Based on Software Transactional Memory on Multicore Architectures

PDP '11 Proceedings of the 2011 19th International Euromicro Conference on Parallel, Distributed and Network-Based Processing
A static task partitioning approach for heterogeneous systems using OpenCL

CC'11/ETAPS'11 Proceedings of the 20th international conference on Compiler construction: part of the joint European conferences on theory and practice of software
A machine learning-based approach for thread mapping on transactional memory applications

HIPC '11 Proceedings of the 2011 18th International Conference on High Performance Computing

Automatic Skeleton-Driven Memory Affinity for Transactional Worklist Applications

International Journal of Parallel Programming

Quantified Score

Hi-index	0.00

Visualization

Abstract

Thread mapping is an appealing approach to efficiently exploit the potential of modern chip-multiprocessors. However, efficient thread mapping relies upon matching the behavior of an application with system characteristics. In particular, Software Transactional Memory (STM) introduces another dimension due to its runtime system support. In this work, we propose a dynamic thread mapping approach to automatically infer a suitable thread mapping strategy for transactional memory applications composed of multiple execution phases with potentially different transactional behavior in each phase. At runtime, it profiles the application at specific periods and consults a decision tree generated by a Machine Learning algorithm to decide if the current thread mapping strategy should be switched to a more adequate one. We implemented this approach in a state-of-the-art STM system, making it transparent to the user. Our results show that the proposed dynamic approach presents performance improvements up to 31% compared to the best static solution.