Pasar al contenido principal

CONSTRUCCION Y OPTIMIZACION AUTOMATICAS DE BIBLIOTECAS PARALELAS DE COMPUTACION CIENTIFICA
- UM

TIN2008-06570-C04-02

•
Nombre agencia financiadora Ministerio de Ciencia e Innovación
Acrónimo agencia financiadora MICINN
Programa Programa Nacional de Proyectos de Investigación Fundamental
Subprograma Subprograma de Proyectos de Investigación Fundamental no Orientada
Convocatoria Proyectos I+D
Año convocatoria 2008
Plan Estatal 2008-2011 Plan Nacional de Investigación Científica, Desarrollo e Innovación Tecnológica
Unidad de gestión Subdirección General de Proyectos de Investigación
Centro beneficiario UNIVERSIDAD DE MURCIA
Identificador persistente http://dx.doi.org/10.13039/501100004837

Resultados relacionados

Resultados totales (Incluyendo duplicados): 10
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Block pivoting implementation of a symmetric Toeplitz solver

  • Alonso-Jordá, Pedro|||0000-0002-6882-6592
  • Vidal Maciá, Antonio Manuel
  • Dolz Zaragozá, Manuel Francisco
Toeplitz matrices are characterized by a special structure that can be exploited in order to obtain fast linear system solvers. These solvers are difficult to parallelize due to their low computational cost and their closely coupled data operations. We propose to transform the Toeplitz system matrix into a Cauchy-like matrix since the latter can be divided into two independent matrices of half the size of the system matrix and each one of these smaller arising matrices can be factorized efficiently in multicore computers. We use OpenMP and store data in memory by blocks in consecutive positions yielding a simple and efficient algorithm. In addition, by exploiting the fact that diagonal pivoting does not destroy the special structure of Cauchy-like matrices, we introduce a local diagonal pivoting technique which improves the accuracy of the solution and the stability of the algorithm., This work was partially supported by the Spanish Ministerio de Ciencia e Innovacion (Project TIN2008-06570-C04-02 and TEC2009-13741), Vicerrectorado de Investigacion de la Universidad Politecnica de Valencia through PAID-05-10 (ref. 2705), and Generalitat Valenciana through project PROMETEO/2009/2013.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Solving systems of symmetric Toeplitz tridiagonal equations: Rojo&apos, s algorithm revisited

  • Vidal Maciá, Antonio Manuel
  • Alonso-Jordá, Pedro|||0000-0002-6882-6592
More than 20 years ago, Rojo published [1] an algorithm for solving linear systems where the matrix is tridiagonal symmetric Toeplitz and diagonal dominant. The technique proposed by Rojo is very efficient, O(n), and has been applied successfully in the solution of other similar problems: circulant tridiagonal systems, pentadiagonal Toeplitz systems, etc. In this article we extend Rojo's algorithm to the case of non-diagonal dominant matrices, thus completing a good tool in the aforementioned applications. Other algorithms that solve the same problem are also analysed and compared with the new version of Rojo's algorithm. © 2012 Elsevier Inc. All rights reserved., Supported by Spanish Government (Projects TIN2008-06570-C04 and TEC2009-13741), and Generalitat Valenciana (Project PROMETEO/2009/013).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Tridimensional block multiword LDPC decoding on GPUs

  • Martínez Zaldívar, Francisco José|||0000-0002-7845-6273
  • Vidal Maciá, Antonio Manuel
  • Gonzalez, Alberto|||0000-0002-6984-3212
  • Almenar Terre, Vicenç|||0000-0002-8742-2315
[EN] In this paper, we describe a parallel algorithm for LDPC (Low Density Parity Check codes) decoding on a GPU (Graphics Processing Unit) using CUDA (Compute Unified Device Architecture). The strategy of the kernel grid and block design is shown and the multiword decoding operation is described using tridimensional blocks. The performance (speedup) of the proposed parallel algorithm is slightly better than the performance found in the literature when this is relatively good, and shows a great improvement in those cases with previously reported moderate or bad performance. © 2011 Springer Science+Business Media, LLC., This work was financially supported by the Spanish Ministerio de Ciencia e Innovación (Projects TIN2008-06570-C04-02, TEC2009-13741 and TEC2008-06787), Universidad Politécnica de Valencia through “Programa de Apoyo a la Investigación y Desarrollo (PAID-05-09)” and Generalitat Valenciana through project PROMETEO/2009/013.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Real-time massive convolution for audio applications on GPU

  • Belloch Rodríguez, José Antonio
  • Gonzalez, Alberto|||0000-0002-6984-3212
  • Martínez Zaldívar, Francisco José|||0000-0002-7845-6273
  • Vidal Maciá, Antonio Manuel
[EN] Massive convolution is the basic operation in multichannel acoustic signal processing. This field has experienced a major development in recent years. One reason for this has been the increase in the number of sound sources used in playback applications available to users. Another reason is the growing need to incorporate new effects and to improve the hearing experience. Massive convolution requires high computing capacity. GPUs offer the possibility of parallelizing these operations. This allows us to obtain the processing result in much shorter time and to free up CPU resources. One important aspect lies in the possibility of overlapping the transfer of data from CPU to GPU and vice versa with the computation, in order to carry out real-time applications. Thus, a synthesis of 3D sound scenes could be achieved with only a peer-to-peer music streaming environment using a simple GPU in your computer, while the CPU in the computer is being used for other tasks. Nowadays, these effects are obtained in theaters or funfairs at a very high cost, requiring a large quantity of resources. Thus, our work focuses on two mains points: to describe an efficient massive convolution implementation and to incorporate this task to real-time multichannel-sound applications. © 2011 Springer Science+Business Media, LLC., This work was partially supported by the Spanish Ministerio de Ciencia e Innovacion (Projects TIN2008-06570-C04-02 and TEC2009-13741), Universidad Politecnica de Valencia through PAID-05-09 and Generalitat Valenciana through project PROMETEO/2009/2013
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

The impact of GPU/Multicore in Signal Processing: a quantitative approach

  • García Mollá, Víctor Manuel|||0000-0003-4768-7367
  • Gonzalez, Alberto|||0000-0002-6984-3212
  • Martínez Zaldívar, Francisco José|||0000-0002-7845-6273
  • Vidal Maciá, Antonio Manuel
  • González García, Cristina Yenixe
  • Ramiro Sánchez, Carla
  • Roger Varea, Sandra
[EN] This paper presents a meaningful practical performance comparison between the last generation of Graphics Processing Units (GPUs) and the last generation multi-core CPUs when they are used to solve given Signal Processing algorithms. Two kinds of tests were considered: when GPU pre-designed computational libraries were available, and when the GPU code was developed by the authors. Results show that GPUs offer great possibilities, but its programming is still hard and high performances can be obtained only if the algorithm can be adapted to the GPU programming model., This work was financially supported by the Spanish Ministerio de Ciencia e Innovación (Projects TIN2008-06570-C04-02, TEC2009-13741 and CAPAP-H3 TIN2010-12011-E), Universitat Politècnica de València through “Programa de Apoyo a la Investigación y Desarrollo (PAID-05-10)” and Generalitat Valenciana through project PROMETEO/2009/013.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Estudio, informe, memoria (Report)

Parallel Two-Stage Least Squares algorithms for Simultaneous Equations Models on GPU

  • Ramiro Sánchez, Carla
  • López-Espín, José J.
  • Giménez, Domingo
  • Vidal, Antonio M.
Today it is usual to have computational systems formed by a multicore together with one or more GPUs. These systems are heterogeneous, due to the di erent types of memory in the GPUs and to the di erent speeds of computation of the cores in the CPU and the GPU. To accelerate the solution of complex problems it is necessary to combine the two basic components (CPU and GPU) in the heterogeneous system. This paper analyzes the use of a multicore+multiGPU system for solving Simultaneous Equations Models by the Two-Stage Least Squares method with QR decomposition. The combination of CPU and GPU allows us to reduce the execution time in the solution of large SEM.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Heterogeneous pipelined square-root Kalman Filter algorithm for the MMSE-OSIC problem

  • Martínez Zaldívar, Francisco José|||0000-0002-7845-6273
  • Vidal Maciá, Antonio Manuel
  • Giménez Cánovas, Domingo
[EN] This paper describes a pipelined parallel algorithm for the MMSE-OSIC decoding procedure proposed in V-BLAST wireless MIMO systems, for heterogeneous networks of processors. It is based on a block version of the square-root Kalman Filter algorithm that was initially devised to solve the RLS problem. It has been parallelized in a pipelined way obtaining a good efficiency and scalability. The optimum load balancing for this parallel algorithm is dynamic, but we derive a static load balancing scheme with good performance. © 2009 Springer Science+Business Media, LLC., This work has been supported by the Generalitat Valenciana, project 20080811, by the Universidad Politecnica de Valencia, project 20080009, by the Conserjería de Educacion de la Región de Murcia (Fundacion Séneca, 08763/PI/08), and by the Ministerio de Ciencia e Innovacion (TIN2008-06570-C04-02).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

An efficient and scalable block parallel algorithm of Neville elimination as a tool for the CMB maps problem

  • Alonso, P.
  • Cortina, R.
  • Ranilla, J.
  • Vidal Maciá, Antonio Manuel
Abstract This paper analyses the performance of several versions of a block parallel algorithm in order to apply Neville elimination in a distributed memory parallel computer. Neville elimination is a procedure to transform a square matrix A into an upper triangular one. This analysis must take into account the algorithm behaviour as far as execution time, efficiency and scalability are concerned. Special attention has been paid to the study of the scalability of the algorithms trying to establish the relationship existing between the size of the block and the performance obtained in this metric. It is important to emphasize the high efficiency achieved in the studied cases and that the experimental results confirm the theoretical approximation. Therefore, we have obtained a high predicting ability tool of analysis. Finally, we will present the elimination of Neville as an efficient tool in detecting point sources in cosmic microwave background maps.© Springer Science+Business Media, LLC 2010, This work has been partially supported by the Spanish Research Grants TIN2007-61273, TIN2008-06570-C04-02 and TIN2010-14971, and by Valencia Regional Government Grant PRO-METEO/2009/013. Also, we would like to give special thanks to Professor Francisco Argueso of University of Oviedo for his help.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

DVFS-control techniques for dense linear algebra operations on multi-core processors

  • Alonso-Jordá, Pedro|||0000-0002-6882-6592
  • Quintana-Ortí, Enrique S.|||0000-0002-5454-165X
  • Dolz Zaragozá, Manuel Francisco
  • Igual, Francisco D.
  • Mayo, Rafael
[EN] This paper analyzes the impact on power con- sumption of two DVFS-control strategies when applied to the execution of dense linear algebra operations on multi- core processors. The strategies considered here, prototyped as the Slack Reduction Algorithm (SRA) and the Race-to- Idle Algorithm (RIA), adjust the operation frequency of the cores during execution of a collection of tasks (in which many dense linear algebra algorithms can be decomposed) with a very different approach to save energy. A power- aware simulator, in charge of scheduling the execution of tasks to processor cores, is employed to evaluate the perfor- mance benefits of these power-control policies for two ref- erence algorithms for the LU factorization, a key operation for the solution of linear systems of equations., The authors from Univ. Jaume I were supported by project CICYT TIN2008-06570-C04 and FEDER.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Detecting point sources in CMB maps using an efficient parallel algorithm

  • Alonso, P.
  • Argüeso, F.
  • Cortina, R.
  • Ranilla, J.
  • Vidal Maciá, Antonio Manuel
The Cosmic Microwave Background (CMB) is a diffuse radiation which is contaminated by the radiation emitted by point sources. The precise knowledge of CMB fluctuations can lead to a better knowledge of the chemistry at the early stages of the Universe. In this work, we present an efficient algorithm, with a high degree of parallelism, which can improve, from the computational point of view, the classical approaches for detecting point sources in Cosmic Microwave Background maps. High performance computing libraries and parallel computing techniques have allowed to construct a portable, fast and numerically stable algorithm. To check the performance of the new method, we have carried out several simulations resembling the observational data collected by the Low Frequency Instrument of the Planck satellite. The sources are detected in their real positions. © 2011 Springer Science+Business Media, LLC., This work was financially supported by the Spanish Ministerio de Ciencia e Innovacion and by FEDER (Projects TIN2010-14971, TIN2008-06570-C04-02, TEC2009-13741 and CAPAP-H3 TIN2010-12011-E), Universitat Politecnica de Valencia through Programa de Apoyo a la Investigacion y Desarrollo (PAID-05-10) and Generalitat Valenciana through project PROMETEO/2009/013. We thank Diego Herranz for his help with the CMB simulations.