Performance Evaluation of Parallel Sparse Matrix–Vector Products on SGI Altix3700

Kotakemori, Hisashi; Hasegawa, Hidehiko; Kajiyama, Tamito; Nukada, Akira; Suda, Reiji; Nishida, Akira

doi:10.1007/978-3-540-68555-5_13

Cited by 17 publications

(12 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This operation is considered as a relevant computational kernel in scientific applications, which performs not optimally on modern processors because of the lack of compromise between memory and computing power and irregular memory access patterns ). In general, we find quite a lot of done work in the field of sparse matrix-vector multiplications using parallelization techniques (Kotakemori et al, 2008;Goumas et al, 2009;Williams et al, 2009). These papers study in depth the optimal performance of this operation, but in this paper, we show that even using a simpler parallelization routine, the computation time is noticeably shortened.…”

Section: Related Workmentioning

confidence: 83%

Parallelization of a Mathematical Model to Evaluate a Cca Application for Vanets

Murcia-Hernández¹,

García-Costa²,

Tomás-Gabarrón³

et al. 2011

Proceedings of 1st International Conference on Simulation and Modeling Methodologies, Technologies and Applications

View full text Add to dashboard Cite

Abstract:Vehicular Ad-hoc Networks (VANET) are currently becoming not only an extremely important factor for vehicles engineering development but also a key issue for improving road safety. Cooperative/Chain Collision Avoidance (CCA) application comes up as a solution for decreasing accidents on the road, therefore it is highly convenient to study how the system of vehicles in a platoon will behave at different stages of technology deployment until full penetration in the market. We have developed an analytical model to compute the average number of accidents in a platoon of vehicles. However, due to the model structure, when the CCA technology penetration rate is taken into account, the increase in the number of operations of the analytical model is such that the sequential computation of a numerical solution is no longer feasible. In this paper, with the goal in mind of reducing computation time, we show how we have implemented and parallelized our analytical model so as a solution can be achieved, what is conducted using the OpenMP parallelization techniques under a supercomputing shared memory environment.

show abstract

Section: Related Workmentioning

confidence: 83%

Parallelization of a Mathematical Model to Evaluate a Cca Application for Vanets

Murcia-Hernández¹,

García-Costa²,

Tomás-Gabarrón³

et al. 2011

Proceedings of 1st International Conference on Simulation and Modeling Methodologies, Technologies and Applications

View full text Add to dashboard Cite

show abstract

“…This operation is considered as a relevant computational kernel in scientific applications, which performs not optimally on modern processors because of the lack of compromise between memory and computing power and irregular memory access patterns [5]. In general, we find quite a lot of done work in the field of sparse matrix-vector multiplications using parallelization techniques [6], [7], [8]. These papers study in depth the optimal performance of this operation, but in this paper, we show that even using a simpler parallelization routine, the computation time is noticeably shortened.…”

Section: Related Workmentioning

confidence: 99%

Speeding Up the Evaluation of a Mathematical Model for VANETs Using OpenMP

García-Costa

Tomás-Gabarrón

Egea-López

et al. 2013

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

Abstract. Vehicular Ad-hoc Networks (VANETs) are having a significant impact on Intelligent Transportation Systems, specially on the improvement of road safety. Cooperative/Chain Collision Avoidance (CCA) application comes up as a solution for decreasing accidents on the road, therefore it is highly convenient to study how the system of vehicles in a platoon will behave at different stages of technology deployment until full penetration in the market. In the present paper we describe an analytical model to compute the average number of accidents in a chain of vehicles. The use of this model when the CCA technology penetration rate is not 100% leads to a vast increase in the number of operations. Using the OpenMP directives for parallel processing with shared memory we achieve a significant reduction in the computation time consumed by our analytical model.

show abstract

“…In [4], the authors analyze the behavior of the SpMV on a NUMA Altix system, focusing on the effect of using different sparse matrix storage formats. Williams et al [5] propose several optimization techniques for the SpMV which are evaluated on different multicore platforms.…”

Section: Introductionmentioning

confidence: 99%

Analyzing the execution of sparse matrix-vector product on the Finisterrae SMP-NUMA system

et al. 2010

View full text Add to dashboard Cite

In this paper, the sparse matrix-vector product (SpMV) is evaluated on the FINISTERRAE SMP-NUMA supercomputer. Its architecture particularities make the tuning of SpMV especially relevant due to the significant impact on the performance. First, we have estimated the influence of data and thread allocation. Moreover, because of the indirect and irregular memory access patterns of SpMV, we have also studied the influence of the memory hierarchy in the performance. According to the behavior observed in the study, a set of optimizations specially tuned for FINISTER-RAE were successfully applied to SpMV. Noticeable improvements are obtained in comparison with the SpMV naïve implementation.

show abstract

Performance Evaluation of Parallel Sparse Matrix–Vector Products on SGI Altix3700

Cited by 17 publications

References 8 publications

Parallelization of a Mathematical Model to Evaluate a Cca Application for Vanets

Parallelization of a Mathematical Model to Evaluate a Cca Application for Vanets

Speeding Up the Evaluation of a Mathematical Model for VANETs Using OpenMP

Analyzing the execution of sparse matrix-vector product on the Finisterrae SMP-NUMA system

Contact Info

Product

Resources

About