Towards the Scalability of Dynamic Loop Scheduling Techniques via Discrete Event Simulation

Balasubramaniam, Mahadevan; Sukhija, Nitin; Ciorba, Florina M.; Banicescu, Ioana; Srivastava, Srishti

doi:10.1109/ipdpsw.2012.171

Cited by 15 publications

(25 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Over the years, many DLS techniques have been studied and well described in the literature [14], [15]. Among them, two categories can be considered: non-adaptive and adaptive.…”

Section: A Dynamic Loop Schedulingmentioning

confidence: 99%

An Autonomic Approach for the Selection of Robust Dynamic Loop Scheduling Techniques

Boulmier

Banicescu

Ciorba

et al. 2017

2017 16th International Symposium on Parallel and Distributed Computing (ISPDC)

Self Cite

View full text Add to dashboard Cite

Abstract-Parallel applications are highly irregular and high performance computing (HPC) infrastructures are very complex. The HPC applications of interest herein are timestepping scientific applications (TSSA). Often, TSSA involve the repeated execution of multiple parallel loops with thousands of iterations and irregular behavior. Dynamic loop scheduling (DLS) techniques were developed over time and have proven to be effective in scheduling parallel loops for achieving load balancing of TSSA. Using a single particular DLS technique throughout the entire execution of a time-step, or even over the entire application, does not guarantee optimal performance due to the unpredictable variations in problem and algorithmic characteristics as well as those of the infrastructure capabilities. For that reason, an autonomic selection of DLS techniques as function of the parallel loop execution time has shown to improve application performance. Recently, a robustness metric of DLS techniques, named "flexibility", has been proposed to estimate the capability of a DLS technique to resist to variations in the loop iterations execution time. To improve the performance of TSSA, we propose in this work an approach that involves the autonomic selection of DLS techniques as function of the flexibility of DLS techniques. The first major novelty of our approach lies in the use of state-of-the-art reinforcement learning (RL) algorithms as smart agents. The second novelty lies in the design of a modified flexibility metric. The third major novelty resides in using the new modified flexibility metric as a reward for the smart agents. The fourth novelty is the evaluation of the proposed approach within a simulated environment, in particular using the SimGrid-SMPI interface to execute DLS algorithms. We discuss the advantages and the limitations of the new proposed flexibility metric as a reward.

show abstract

“…Over the years, many DLS techniques have been studied and well described in the literature [14], [15]. Among them, two categories can be considered: non-adaptive and adaptive.…”

Section: A Dynamic Loop Schedulingmentioning

confidence: 99%

An Autonomic Approach for the Selection of Robust Dynamic Loop Scheduling Techniques

Boulmier

Banicescu

Ciorba

et al. 2017

2017 16th International Symposium on Parallel and Distributed Computing (ISPDC)

Self Cite

View full text Add to dashboard Cite

show abstract

“…First, in contrast to the related works that propose a simulation-based methodology for evaluating loop scheduling strategies [6,13], we propose a methodology that not only enables the evaluation of such strategies, but also the design of new ones. First, in contrast to the related works that propose a simulation-based methodology for evaluating loop scheduling strategies [6,13], we propose a methodology that not only enables the evaluation of such strategies, but also the design of new ones.…”

Section: Related Workmentioning

confidence: 99%

“…Other strategies were not considered because the dynamic one results in similar performance gains when an optimal chunk size is used [13]. Other strategies were not considered because the dynamic one results in similar performance gains when an optimal chunk size is used [13].…”

Section: Experimental Evaluationmentioning

confidence: 99%

Design methodology for workload‐aware loop scheduling strategies based on genetic algorithm and simulation

Penna

Castro

Freitas

et al. 2016

Concurrency and Computation

View full text Add to dashboard Cite

International audienceIn high-performance computing, the application's workload must be evenly balanced among threads to deliver cutting-edge performance and scalability. In OpenMP, the load balancing problem arises when scheduling loop iterations to threads. In this context, several scheduling strategies have been proposed, but they do not take into account the input workload of the application and thus turn out to be suboptimal. In this work, we introduce a design methodology to propose, study, and assess the performance of workload-aware loop scheduling strategies. In this methodology, a genetic algorithm is employed to explore the state space solution of the problem itself and to guide the design of new loop scheduling strategies, and a simulator is used to evaluate their performance. As a proof of concept, we show how the proposed methodology was used to propose and study a new workload-aware loop scheduling strategy named smart round-robin (SRR). We implemented this strategy into GNU Compiler Collection's OpenMP runtime. We carry out several experiments to validate the simulator and to evaluate the performance of SRR. Our experimental results show that SRR may deliver up to 37.89% and 14.10% better performance than OpenMP's dynamic loop scheduling strategy in the simulated environment and in a real-world application kernel, respectively

show abstract

“…Balasubramaniam et al [2] and Byung et al [14] presented aproaches targeting Desktop Grids, and Godfrey et al [4], Shudo [13], Senís et al [12] and Wu and Tian [15] present aproaches targeting PDG.…”

Section: Related Workmentioning

confidence: 99%

Rescheduling and checkpointing as strategies to run synchronous parallel programs on P2P desktop grids

Righi

Veith

Rodrigues

et al. 2015

Proceedings of the 30th Annual ACM Symposium on Applied Computing

View full text Add to dashboard Cite

Today, BSP (Bulk-Synchronous Parallel) represents one of the most often used models for writing tightly-coupled parallel programs. As resource substrates, commonly clusters and eventually computational grids are used to run BSP applications. In this context, here we investigate the use of collaborative computing and idle resources to execute this kind of demand, so we are proposing a model named BSPonP2P to answer the following question: How can we develop an e cient and viable model to run BSP applications on P2P Desktop Grids? We answer it by providing both process rescheduling and checkpointing to deal with dynamism at application and infrastructure levels and resource heterogeneity. The results concern a prototype that ran over a subset of the Grid5000, showing encouraging results on using collaboration and volatile resources for HPC.

show abstract

Towards the Scalability of Dynamic Loop Scheduling Techniques via Discrete Event Simulation

Cited by 15 publications

References 26 publications

An Autonomic Approach for the Selection of Robust Dynamic Loop Scheduling Techniques

An Autonomic Approach for the Selection of Robust Dynamic Loop Scheduling Techniques

Design methodology for workload‐aware loop scheduling strategies based on genetic algorithm and simulation

Rescheduling and checkpointing as strategies to run synchronous parallel programs on P2P desktop grids

Contact Info

Product

Resources

About