Advanced search options

Advanced Search Options 🞨

Browse by author name (“Author name starts with…”).

Find ETDs with:

in
/  
in
/  
in
/  
in

Written in Published in Earliest date Latest date

Sorted by

Results per page:

Sorted by: relevance · author · university · dateNew search

You searched for subject:(Processadores multicore). Showing records 1 – 2 of 2 total matches.

Search Limiters

Last 2 Years | English Only

No search limiters apply to these results.

▼ Search Limiters

1. Leonardo Garcia Marques. Programação genética paralela com Pareto: uma ferramenta para modelagem via regressão simbólica.

Degree: 2013, Federal University of Uberlândia

Indução de programas envolve a descoberta de programas de computador que produzem alguma saída desejada quando estes são submetidos a alguma entrada em particular. Um exemplo é a regressão simbólica, ferramenta de modelagem que busca expressões de funções matemáticas para ajustar determinado conjunto de dados multivariados, mapeando variáveis de entrada para variáveis de saída de controle. A programação genética, uma sub-área da computação evolutiva que usa analogia da teoria da evolução de Darwin e algumas ideias de genética, é uma técnica automática para produzir programas de computador amplamente usada para resolver problemas. No entanto, a implementação da programação genética não é trivial para a maioria dos profissionais, além de demandar alto poder computacional. Este trabalho apresenta uma implementação paralela de programação genética simples de se manusear, otimizada para computadores de arquitetura com múltiplos núcleos e que satisfaz o critério competitivo de simplicidade estrutural e exatidão na predição, através de variação especial multiobjetiva de programação genética, chamada programação genética com Pareto. A implementação proposta tem ganhos de desempenho proporcionais à quantidade de núcleos disponíveis em uso, além de ter sido aplicada com sucesso em diversos tipos de problemas de regressão.

Program induction involves the inductive discovery of a computer program that produces some desired output when presented with some particular input. An example is the symbolic regression, a modeling tool that seeks mathematical expressions of functions to fit a given multivariate data set, mapping input variables to output variables of control. The genetic programming, a subarea of evolutive computing that uses an analogy of Darwins evolutionary theory and some ideas from the genetics field, is an automatic technique for producing a computer program widely used to solve such problems. However, implementing genetic programming is not trivial for most professionals, besides demanding high computational power. This work presents a parallel implementation of genetic programming simple to handle, optimized for computers with multicore architecture, and satisfying competitive criteria of structural simplicity model and prediction accurate model, through a special multi-objective flavor of a genetic programming, called Pareto Genetic Programing. The proposed implementation has performance gains proportional to the amount of available cores in use, and has been successfully applied to several types of regression problems.

Advisors/Committee Members: Keiji Yamanaka, Sergio A. A. de Freitas, Alexsandro Santos Soares, Wesley Pacheco Calixto.

Subjects/Keywords: Programação genética; Processadores multicore; Dominância de Pareto; ENGENHARIA ELETRICA; Informática; Programação paralela (Computação); Inteligência artificial; Programação genética (Computação); Genetic Programming; Multicore processors; Pareto dominance

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Marques, L. G. (2013). Programação genética paralela com Pareto: uma ferramenta para modelagem via regressão simbólica. (Thesis). Federal University of Uberlândia. Retrieved from http://www.bdtd.ufu.br//tde_busca/arquivo.php?codArquivo=5409

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Marques, Leonardo Garcia. “Programação genética paralela com Pareto: uma ferramenta para modelagem via regressão simbólica.” 2013. Thesis, Federal University of Uberlândia. Accessed September 21, 2020. http://www.bdtd.ufu.br//tde_busca/arquivo.php?codArquivo=5409.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Marques, Leonardo Garcia. “Programação genética paralela com Pareto: uma ferramenta para modelagem via regressão simbólica.” 2013. Web. 21 Sep 2020.

Vancouver:

Marques LG. Programação genética paralela com Pareto: uma ferramenta para modelagem via regressão simbólica. [Internet] [Thesis]. Federal University of Uberlândia; 2013. [cited 2020 Sep 21]. Available from: http://www.bdtd.ufu.br//tde_busca/arquivo.php?codArquivo=5409.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Marques LG. Programação genética paralela com Pareto: uma ferramenta para modelagem via regressão simbólica. [Thesis]. Federal University of Uberlândia; 2013. Available from: http://www.bdtd.ufu.br//tde_busca/arquivo.php?codArquivo=5409

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

2. Lucas, Divino César Soares, 1985-. The batched DOACROSS algorithm = O algoritmo batched DOACROSS: O algoritmo batched DOACROSS.

Degree: 2017, Universidade Estadual de Campinas

Abstract: Parallelizing loops containing loop-carried dependences has been considered a very difficult task, mainly due to the overhead imposed by communicating dependences between iterations. Despite the huge efforts in the past few decades to devise effective parallelization algorithms for such loops, the problem is still far from solved. For many loops, old DOACROSS, and new Decoupled Software Pipeline (DSWP), algorithms have not been able to offer a solution to this problem. This thesis discuss in detail two of the most prominent algorithms for parallelizing such loops and also present an analysis of the performance of the parallelized programs across different multicore architectures. Based on insights from this analyze a new algorithm, called Batched DOACROSS, for parallelizing these loops is proposed. Batched DOACROSS (BDX) capitalizes on the advantages of DSWP and DOACROSS, while minimizing their deficiencies. BDX does not require new hardware mechanisms (as DSWP does) and makes use of thread local buffers to reduce DOACROSS synchronization overheads. An extension to the baseline algorithm is proposed, named Parallel-Stage BDX (PS-BDX), and show that in some cases it can considerably improve the performance of the parallel loop. BDX and PS-BDX are pipelining multithreading algorithms that employs batching to amortize communication overheads. We provide results for a sensibility analysis and show that for small balanced loops (about 40 instructions), a batch size of only 100 iterations is sufficient to provide good speedups. Our analyze of PS-BDX for seven benchmarks showed an average of 1.85x speedup for 2 threads, 2.95x for 4 threads and 3.11x for 8 threads which was larger than the other best algorithm that we compared. A qualitative and quantitative analysis of synchronization costs of the three aforementioned loop parallelization algorithms (BDX, DOACROSS and DSWP) is performed for two modern computer architectures (ARM A9 MPCore and Intel Ivy Bridge). Our results show that at least 30% of the execution time of the programs we parallelized are spent on synchronization/data communication. We also show that, besides the problem being hard, Intel Ivy Bridge and ARM A9 MPCore are on opposite endpoints along the axis of commonly accepted requisites for efficient loop parallelization. As a consequence, all three algorithms struggle to effectively speedup several programs Advisors/Committee Members: UNIVERSIDADE ESTADUAL DE CAMPINAS (CRUESP), Araújo, Guido Costa Souza de, 1962- (advisor), Universidade Estadual de Campinas. Instituto de Computação (institution), Programa de Pós-Graduação em Ciência da Computação (nameofprogram), Baldassin, Alexandro José (committee member), Pereira, Fernando Magno Quintão (committee member), Rigo, Sandro (committee member), Pereira, Marcio Machado (committee member).

Subjects/Keywords: Processamento paralelo (Computadores); Programação paralela (Computação); Processadores multicore; Algoritmos paralelos; Compiladores (Computadores); Parallel processing (Electronic computers); Parallel programming (Computer science); Multicore processors; Parallel algorithms; Compiling (Electronic computers)

…Parallelism (ILP) [1, 3, 48, 69]. The advent of multicore and Simultaneous… …multicore architectures together with the decrease in clock frequency of individual processors… …multicore architectures and state-of-the-art algorithms for loop parallelization. It proposes a… …new algorithm for efficient automatic parallelization of DOACROSS loops for multicore… …and fine granularity across different modern multicore architectures. Furthermore, we also… 

Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Lucas, Divino César Soares, 1. (2017). The batched DOACROSS algorithm = O algoritmo batched DOACROSS: O algoritmo batched DOACROSS. (Thesis). Universidade Estadual de Campinas. Retrieved from http://repositorio.unicamp.br/jspui/handle/REPOSIP/330850

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Lucas, Divino César Soares, 1985-. “The batched DOACROSS algorithm = O algoritmo batched DOACROSS: O algoritmo batched DOACROSS.” 2017. Thesis, Universidade Estadual de Campinas. Accessed September 21, 2020. http://repositorio.unicamp.br/jspui/handle/REPOSIP/330850.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Lucas, Divino César Soares, 1985-. “The batched DOACROSS algorithm = O algoritmo batched DOACROSS: O algoritmo batched DOACROSS.” 2017. Web. 21 Sep 2020.

Vancouver:

Lucas, Divino César Soares 1. The batched DOACROSS algorithm = O algoritmo batched DOACROSS: O algoritmo batched DOACROSS. [Internet] [Thesis]. Universidade Estadual de Campinas; 2017. [cited 2020 Sep 21]. Available from: http://repositorio.unicamp.br/jspui/handle/REPOSIP/330850.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Lucas, Divino César Soares 1. The batched DOACROSS algorithm = O algoritmo batched DOACROSS: O algoritmo batched DOACROSS. [Thesis]. Universidade Estadual de Campinas; 2017. Available from: http://repositorio.unicamp.br/jspui/handle/REPOSIP/330850

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

.