Search Results

You are looking at 1 - 5 of 5 items for

  • Author: J. Blazewicz x
Clear All Modify Search
Open access

J. Nawrocki, W. Complak, J. Błażewicz, S. Kopczyńska and M. Maćkowiaki

The Knapsack-Lightening problem and its application to scheduling HRT tasks

In hard real-time systems timeliness is as important as functional correctness. Such systems contain so called hard real-time tasks (HRT tasks) which must be finished by a given deadline. One of the methods of scheduling of HRT tasks is periodic loading introduced by Schweitzer, Dror, and Trudeau. The paper presents an extension to that method which allows for deterministic utilization of cache memory in hard real-time systems. It is based on a new version of the Knapsack problem named Knapsack-Lightening. In the paper the Knapsack-Lightening problem is defined, its complexity is analyzed, and an exact algorithm along with two heuristics are presented. Moreover the application of the Knapsack-Lightening problem to scheduling HRT tasks is described.

Open access

W. Frohmberg, M. Kierzynka, J. Blazewicz and P. Wojciechowski

Abstract

Several highly efficient alignment tools have been released over the past few years, including those taking advantage of GPUs (Graphics Processing Units). G-PAS (GPU-based Pairwise Alignment Software) was one of them, however, with a couple of interesting features that made it unique. Nevertheless, in order to adapt it to a new computational architecture some changes had to be introduced. In this paper we present G-PAS 2.0 - a new version of the software for performing high-throughput alignment. Results show, that the new version is faster nearly by a fourth on the same hardware, reaching over 20 GCUPS (Giga Cell Updates Per Second).

Open access

W. Frohmberg, M. Kierzynka, J. Blazewicz, P. Gawron and P. Wojciechowski

Abstract

DNA/RNA sequencing has recently become a primary way researchers generate biological data for further analysis. Assembling algorithms are an integral part of this process. However, some of them require pairwise alignment to be applied to a great deal of reads. Although several efficient alignment tools have been released over the past few years, including those taking advantage of GPUs (Graphics Processing Units), none of them directly targets high-throughput sequencing data. As a result, a need arose to create software that could handle such data as effectively as possible. G-DNA (GPU-based DNA aligner) is the first highly parallel solution that has been optimized to process nucleotide reads (DNA/RNA) from modern sequencing machines. Results show that the software reaches up to 89 GCUPS (Giga Cell Updates Per Second) on a single GPU and as a result it is the fastest tool in its class. Moreover, it scales up well on multiple GPUs systems, including MPI-based computational clusters, where its performance is counted in TCUPS (Tera CUPS).

Open access

J. Musial, J.E. Pecero, M.C. Lopez-Loces, H.J. Fraire-Huacuja, P. Bouvry and J. Blazewicz

Abstract

The Internet shopping optimization problem arises when a customer aims to purchase a list of goods from a set of web-stores with a minimum total cost. This problem is NP-hard in the strong sense. We are interested in solving the Internet shopping optimization problem with additional delivery costs associated to the web-stores where the goods are bought. It is of interest to extend the model including price discounts of goods.

The aim of this paper is to present a set of optimization algorithms to solve the problem. Our purpose is to find a compromise solution between computational time and results close to the optimum value. The performance of the set of algorithms is evaluated through simulations using real world data collected from 32 web-stores. The quality of the results provided by the set of algorithms is compared to the optimal solutions for small-size instances of the problem. The optimization algorithms are also evaluated regarding scalability when the size of the instances increases. The set of results revealed that the algorithms are able to compute good quality solutions close to the optimum in a reasonable time with very good scalability demonstrating their practicability.

Open access

Mario C. Lopez-Loces, Jedrzej Musial, Johnatan E. Pecero, Hector J. Fraire-Huacuja, Jacek Blazewicz and Pascal Bouvry

Abstract

Internet shopping has been one of the most common online activities, carried out by millions of users every day. As the number of available offers grows, the difficulty in getting the best one among all the shops increases as well. In this paper we propose an integer linear programming (ILP) model and two heuristic solutions, the MinMin algorithm and the cellular processing algorithm, to tackle the Internet shopping optimization problem with delivery costs. The obtained results improve those achieved by the state-of-the-art heuristics, and for small real case scenarios ILP delivers exact solutions in a reasonable amount of time.