Table of contents

The Program committee of HPCS2013 would like to thank those who contributed to HPCS2013, through the technical program, the Birds of Feather sessions, the vendor overviews, the networking sessions, or for attending and grilling the speakers in all of theses sessions with great questions and contributing to fantastic discussions.

We'd particularly like to highlight the best paper award presented at the conference, going to ''The Making of Big Brain'', presented by Marc-Étienne Rousseau for the Big Brain team; the best student paper for ''Towards a Resource Reservation Approach for an Opportunistic Computing Environment'', presented by Eliza Gomes; and the best visualization, to a movie of an amazing globe-to-individual-building level simulation of the evolution of a toxic plume over a city, presented by Bertrand Denis of the Canadian Meteorological Centre.

It was a great conference, and we look forward to seeing you in Halifax for HPCS2014!

All papers published in this volume of Journal of Physics: Conference Series have been peer reviewed through processes administered by the proceedings Editors. Reviews were conducted by expert referees to the professional and scientific standards expected of a proceedings journal published by IOP Publishing.

A well-known problem faced by any organization nowadays is the high volume of data that is available and the required process to transform this volume into differential information. In this study, a case-comparison study of automatic document classification (ADC) approach is presented, utilizing both serial and parallel paradigms. The serial approach was implemented by adopting the RapidMiner software tool, which is recognized as the worldleading open-source system for data mining. On the other hand, considering the MapReduce programming model, the Hadoop software environment has been used. The main goal of this case-comparison study is to exploit differences between these two paradigms, especially when large volumes of data such as Web text documents are utilized to build a category database. In the literature, many studies point out that distributed processing in unstructured documents have been yielding efficient results in utilizing Hadoop. Results from our research indicate a threshold to such efficiency.

Advanced reservation has been used in grid environments to provide quality of service (QoS) and to guarantee resources available at the execution time. However, in grid subtypes, such as opportunistic grid computing, it is a challenge provides QoS and guarantee of availability resources. In this article, we propose a new advanced reservation approach which offers to a user the possibility to select resources in advance for a future utilization. Therefore, the main goal of this proposal is to offer a best effort feature to a user from an opportunistic configuration. In these types of environments, it is not possible to provide QoS, because, usually, there are no guarantees of resources availability and, consequently, the execution of users applications. In addition, this research work provides a way to organize executions, what it can improve the scheduling and system operations. Experimental results, carried out through a case study, shown the efficiency and relevance of our proposal.

In multi-data center computing, data to be processed is not always local to the computation. This is a major challenge especially for data-intensive Cloud computing applications, since large amount of data would need to be either moved the local sites (staging) or accessed remotely over the network (remote I/O). Cloud application developers generally chose between staging and remote I/O intuitively without making any scientific comparison specific to their application data access patterns since there is no generic model available that they can use. In this paper, we propose a generic model for the Cloud application developers which would help them to choose the most appropriate data access mechanism for their specific application workloads. We define the parameters that potentially affect the end-to-end performance of the multi-data center Cloud applications which need to access large datasets over the network. To test and validate our models, we implemented a series of synthetic benchmark applications to simulate the most common data access patterns encountered in Cloud applications. We show that our model provides promising results in different settings with different parameters, such as network bandwidth, server and client capabilities, and data access ratio.

In recent years, large scale computer systems have emerged to meet the demands of high storage, supercomputing, and applications using very large data sets. The emergence of Cloud Computing offers the potentiel for analysis and processing of large data sets.

Mapreduce is the most popular programming model which is used to support the development of such applications. It was initially designed by Google for building large datacenters on a large scale, to provide Web search services with rapid response and high availability.

In this paper we will test the clustering algorithm K-means Clustering in a Cloud Computing. This algorithm is implemented on MapReduce. It has been chosen for its characteristics that are representative of many iterative data analysis algorithms. Then, we modify the framework CloudSim to simulate the MapReduce execution of K-means Clustering on different Cloud Computing, depending on their size and characteristics of target platforms.

The experiment show that the implementation of K-means Clustering gives good results especially for large data set and the Cloud infrastructure has an influence on these results.

Cloud configurations can be computational environment with interesting cost efficiency for several organizations sizes. However, the indiscriminate action of buying servers and network devices may not represent a correspondent performance number. In the academic and commercial literature, some researches highlight that these environments are idle for long periods. Therefore, energy management is an essential approach in any organization, because energy bills can causes remarkable negative impacts to these organizations in term of costs. In this paper, we present a research work that is characterized by an analysis of energy consumption in a private cloud computing environment, considering both computational resources and network devices. This study was motivated by a real case of a large organization. Therefore, the first part of the study we considered empirical experiments. In a second moment we used the GreenCloud simulator which was utilized to foresee some different configurations. The research reached a successful and differentiated goal in presenting key issues from computational resources and network, related to the energy consumption for real private cloud.

Recently, an alternative strategy for the parallelization of molecular dynamics simulations with short-ranged forces has been proposed. In this work, this algorithm is tested on a variety of multi-core systems using three types of benchmark simulations. The results show that the new algorithm gives consistent speedups which are depending on the properties of the simulated system either comparable or superior to those obtained with spatial decomposition. Comparisons of the parallel speedup on different systems indicates that on multi-core machines the parallel efficiency of the method is mainly limited by memory access speed.

Much work has recently been reported in parallel GPU-based particle swarm optimization (PSO). Motivated by the encouraging results of these investigations, while also recognizing the limitations of GPU-based methods for big problems using a large amount of data, this paper explores the efficacy of employing other types of parallel hardware for PSO. Most commodity systems feature a variety of architectures whose high-performance capabilities can be exploited. In this paper, high-dimensional problems and those that employ a large amount of external data are explored within the context of heterogeneous systems. Large problems are decomposed into constituent components, and analyses are undertaken of which components would benefit from multi-core or GPU parallelism. The current study therefore provides another demonstration that "supercomputing on a budget" is possible when subtasks of large problems are run on hardware most suited to these tasks. Experimental results show that large speedups can be achieved on high dimensional, data-intensive problems. Cost functions must first be analysed for parallelization opportunities, and assigned hardware based on the particular task.

Petroleum reservoir engineering is a complex and interesting field that requires large amount of computational facilities to achieve successful results. Usually, software environments for this field are developed without taking care out of possible interactions and extensibilities required by reservoir engineers. In this paper, we present a research work which it is characterized by the design and implementation based on a software product line model for a real distributed reservoir engineering environment. Experimental results indicate successfully the utilization of this approach for the design of distributed software architecture. In addition, all components from the proposal provided greater visibility of the organization and processes for the reservoir engineers.