High Performance Computing

Esteban Mocskos Sergio Nesmachnow (Eds.) Communications in Computer and Information Science High Performance Computing 4th Latin American Conference, CARLA 2017 Buenos Aires, Argentina, and Colonia del Sacramento, Uruguay, September 20–22, 2017 Revised Selected Papers 123 796 Communications in Computer and Information Science Commenced Publication in 2007 Founding and Former Series Editors: Alfredo Cuzzocrea, Xiaoyong Du, Orhun Kara, Ting Liu, Dominik Ślęzak, and Xiaokang Yang Editorial Board Simone Diniz Junqueira Barbosa Pontifical Catholic University of Rio de Janeiro (PUC-Rio), Rio de Janeiro, Brazil Phoebe Chen La Trobe University, Melbourne, Australia Joaquim Filipe Polytechnic Institute of Setúbal, Setúbal, Portugal Igor Kotenko St Petersburg Institute for Informatics and Automation of the Russian Academy of Sciences, St Petersburg, Russia Krishna M Sivalingam Indian Institute of Technology Madras, Chennai, India Takashi Washio Osaka University, Osaka, Japan Junsong Yuan Nanyang Technological University, Singapore, Singapore Lizhu Zhou Tsinghua University, Beijing, China 796 More information about this series at http://www.springer.com/series/7899 Esteban Mocskos Sergio Nesmachnow (Eds.) • High Performance Computing 4th Latin American Conference, CARLA 2017 Buenos Aires, Argentina, and Colonia del Sacramento, Uruguay, September 20–22, 2017 Revised Selected Papers 123 Editors Esteban Mocskos CSC-CONICET and Universidad de Buenos Aires Buenos Aires Argentina Sergio Nesmachnow Universidad de la República Montevideo Uruguay ISSN 1865-0929 ISSN 1865-0937 (electronic) Communications in Computer and Information Science ISBN 978-3-319-73352-4 ISBN 978-3-319-73353-1 (eBook) https://doi.org/10.1007/978-3-319-73353-1 Library of Congress Control Number: 2017963753 © Springer International Publishing AG 2018 This work is subject to copyright All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed The use of general descriptive names, registered names, trademarks, service marks, etc in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations Printed on acid-free paper This Springer imprint is published by Springer Nature The registered company is Springer International Publishing AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Preface CARLA 2017 September 20-22, 2017 Colonia, Uruguay & Buenos Aires, Argentina High-performance computing (HPC) is a dynamic field that combines the use of innovative computing technologies and algorithms with advances in a broad range of scientific, technical, and industrial areas Latin America shares the global enthusiasm embracing and pushing forward HPC New challenges coming from the use of the computing capabilities of massive multicores, accelerators, cluster platforms, cloud federations, and the new perspectives of Internet of Things resources all help to promote the research and innovation in this area Building on the success of the previous editions of CARLA, the High-Performance Computing Latin America Conference (and former HPCLATAM and CLCAR Conferences), the tenth edition was organized in Buenos Aires and Colonia jointly by Universidad de Buenos Aires (Argentina) and Universidad de la República (Uruguay) The main goal of the CARLA conference is to provide a forum fostering the growth of the HPC community in Latin America, through the exchange and dissemination of new ideas, techniques, and research in HPC In 2017, CARLA featured invited talks from academia and industry, full-paper sessions presenting mature work, and new ideas in research and industrial applications including: distributed systems, parallel algorithms and concurrency; GPU and MIC computing; mobile, grid, and cloud computing; big data, data management, and visualization; scientific computing applications; architecture, infrastructure, and HPC data center; HPC computing education and outreach; industrial solutions Satellite events co-located with CARLA 2017 included the meeting of the Cloud Computing for Smart-City Energy Management (CC-SEM STIC-AmSud) Project, Red SCALAC (Servicios de Computación Avanzada para Latinoamérica y el Caribe), and RICAP CYTED (Red Iberoamericana de Computación de Altas Prestaciones) More than 100 researchers, students, technicians, practitioners, and representatives of industry, technology, and state companies and organizations (from more than 20 countries in Latin America, Europe, Asia, and Oceania) attended the event This book introduces the top contributions presented at CARLA 2017, covering all the aforementioned topics As organizers, we think the articles are valuable contributions to the development of HPC in Latin America December 2017 Esteban Mocskos Sergio Nesmachnow Organization Program Chairs Gregoire Danoy Ricardo Medel Esteban Meneses Esteban Mocskos Sergio Nesmachnow Markus Rampp Carlos Sarraute Luiz Angelo Steffenel Mariano Vazquez University of Luxembourg, Luxembourg Ascentio Technologies S.A., Argentina Costa Rica National High Technology Center, Costa Rica CSC-CONICET and Universidad de Buenos Aires, Argentina Universidad de la República, Uruguay Max Planck Computing and Data Facility, Germany Grandata Labs, Argentina Université de Reims Champagne-Ardenne, France Barcelona Supercomputing Center, Spain Program Committee José Pedro Aguerre Hartwig Anzt Carlos J Barrios Leonardo Bautista Gomez Carlos Bederián Pascal Bouvry Carlos Buil Harold Castro Marcio Castro Maria Clicia Stelling de Castro Gerson Cavalheiro Germán Ceballos Andrea Charão Esteban Clua Flavio Colavecchia Daniel Cordeiro Carlos Couder-Castañeda Alvaro Coutinho Adrián Cristal Gregoire Danoy Alvaro de la Ossa Cristian Mateos Diaz Universidad de la República, Uruguay Karlsruhe Institute of Technology, Germany Universidad Industrial de Santander, Colombia Centro Nacional de Supercomputación, Spain CONICET, Argentina University of Luxembourg, Luxembourg Universidad Técnica Federico Santa María, Chile Universidad de los Andes, Colombia Federal University of Santa Catarina (UFSC), Brazil Universidade Estado Rio de Janeiro, Brazil Universidade Federal de Pelotas, Brazil Uppsala University, Sweden Universidad Federal Santa María, Brazil Universidade Federal Fluminense, Brazil Centro Atómico Bariloche, Comisión Nacional de Energía Atómica, Argentina Universidade de São Paulo, Brazil Instituto Politecnico Nacional, Mexico Federal University of Rio de Janeiro, Brazil Barcelona Supercomputing Centre, Spain University of Luxembourg, Luxembourg Universidad de Costa Rica, Costa Rica ISISTAN-CONICET, Universidad Nacional del Centro, Argentina VIII Organization Gilberto Diaz Mario Jose Diván Bernabe Dorronsoro Ernesto Dufrechou Nicolás Erdödy Eduardo Fernandez Ezequiel Ferrero Alejandro Flores-Méndez Emilio Francesquini Joao Gazolla Veronica Gil-Costa Isidoro Gitler Brice Goglin Leo Gonzalez José Luis Gordillo Jesus Cruz Guzman Elisa Heymann Javier Iparraguirre Santiago Iturriaga Salma Jalife Roberto Leon Francisco Luna Renzo Massobrio Rafael Mayo-Garcia Ricardo Medel Esteban Meneses Renato Miceli Barton Miller Esteban Mocskos Philippe Navaux Sergio Nesmachnow Carla Osthoff Alejandro Otero Horacio Paggi Jairo Panetta Claudio J Paz Martín Pedemonte Tomas Perez-Acle Laercio Lima Pilla Javier Principe Universidad Industrial de Santander, Colombia Universidad Nacional de La Pampa, Argentina Universidad de Cadiz, Spain Universidad de la República, Uruguay Open Parallel, New Zealand Universidad de la República, Uruguay Department of Physics and Center for Complexity and Biosystems, Italy CINVESTAV, Mexico University of Campinas, Brazil Universidade Federal Fluminense, Brazil Universidad Nacional San Luis, Argentina ABACUS-CINVESTAV, Mexico Inria, France Universidad Politecnica de Madrid, Spain Universidad Nacional Autónoma de México, México Universidad Nacional Autónoma de México, México Universitat Autịnoma de Barcelona, Spain Universidad Tecnológica Nacional, Argentina Universidad de la República, Uruguay Corporación Universitaria para el Desarrollo de Internet A.C., Mexico Universidad Andres Bello, Chile Universidad de Málaga, Spain Universidad de la República, Uruguay Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas, Spain Ascentio Technologies S.A., Argentina Costa Rica National High Technology Center, Costa Rica CENAI-CIMATEC, Brazil University of Wisconsin-Madison, USA CSC-CONICET and Universidad de Buenos Aires, Argentina Universidade Federal de Rio Grande Sul, Brazil Universidad de la República, Uruguay National Laboratory for Scientific Computing, Brazil Universidad de Buenos Aires and CSC-CONICET, Argentina Universidad Politécnica de Madrid, Spain Instituto Tecnológico de Aeronáutica, Brazil UTN FRC, Argentina Universidad de la República, Uruguay Universidad de Chile, Chile Universidade Federal de Santa Catarina, Brazil Universidad Politecnica de Cataluña - CIMNE, Spain Organization Juan Manuel Ramírez Markus Rampp Vinod Rebello Genghis Ríos Pablo Rodríguez-Bocca Isaac Rudomin Afonso Sales Carlos Sarraute Lucas Mello Schnorr Hermes Senger Alejandro Soba Roberto Souto Luiz Angelo Steffenel Mario Storti Claude Tadonki Gonzalo Tancredi Andrei Tchernykh Jamal Toutouh Tram Truong-Huu Manuel Ujaldón Gabriel Usera Mariano Vazquez Jesus Verduzco Pablo Javier Vidal Nicolás Wolovick Jesús Xamán Alejandro Zunino IX Universidad de Colima, Mexico Max Planck Computing and Data Facility, Germany Universidade Federal Fluminense, Brazil Pontificia Universidad Católica del Perú, Perú Universidad de la República, Uruguay Barcelona Supercomputing Center, Spain Pontificia Universidad Católica Rio Grande Sul, Brazil Grandata Labs, Argentina Universidade Federal Rio Grande Sul, Brazil Universidade Federal de São Carlos, Brazil Comisión Nacional de Energớa Atúmica, Argentina Laboratúrio Nacional de Computaỗóo, Brazil Universitộ de Reims Champagne-Ardenne, France Universidad Nacional del Litoral and CIMEC-CONICET, Argentina MINES ParisTech, PSL, France Universidad de la República, Uruguay Centro de Investigación Científica y de Educación Superior de Ensenada, Mexico Universidad de Málaga, Spain National University of Singapore, Singapore Universidad de Málaga, Spain Universidad de la República, Uruguay Barcelona Supercomputing Center, Spain Instituto Tecnológico de Colima, México Universidad de la Patagonia Austral, Argentina Universidad Nacional de Córdoba, Argentina Centro Nacional de Investigación y Desarrollo Tecnológico, México ISISTAN-CONICET, Argentina Contents HPC Infrastructures and Datacenters A Deep Learning Mapper (DLM) for Scheduling on Heterogeneous Systems Daniel Nemirovsky, Tugberk Arkose, Nikola Markovic, Mario Nemirovsky, Osman Unsal, Adrian Cristal, and Mateo Valero Power Consumption Characterization of Synthetic Benchmarks in Multicores Jonathan Muraña, Sergio Nesmachnow, Santiago Iturriaga, and Andrei Tchernykh Initial Experiences from TUPAC Supercomputer David Vinazza, Alejandro Otero, Alejandro Soba, and Esteban Mocskos 21 38 HPC Industry and Education romeoLAB: A High Performance Training Platform for HPC, GPU and DeepLearning Arnaud Renard, Jean-Matthieu Etancelin, and Michael Krajecki 55 GPU, Multicores, Accelerators Analysis and Characterization of GPU Benchmarks for Kernel Concurrency Efficiency Pablo Carvalho, Lúcia M A Drummond, Cristiana Bentes, Esteban Clua, Edson Cataldo, and Leandro A J Marzulo Parallel Batch Self-Organizing Map on Graphics Processing Unit Using CUDA Habib Daneshpajouh, Pierre Delisle, Jean-Charles Boisson, Michael Krajecki, and Nordin Zakaria Performance Prediction of Acoustic Wave Numerical Kernel on Intel Xeon Phi Processor Víctor Martínez, Matheus Serpa, Fabrice Dupros, Edson L Padoin, and Philippe Navaux 71 87 101 ... 20-22, 2017 Colonia, Uruguay & Buenos Aires, Argentina High- performance computing (HPC) is a dynamic field that combines the use of innovative computing technologies and algorithms with advances... and MIC computing; mobile, grid, and cloud computing; big data, data management, and visualization; scientific computing applications; architecture, infrastructure, and HPC data center; HPC computing. .. multicore computing systems The study aims at providing an insight of the main power consumption characteristics of different applications when executing over current high performance computing