Implementing the Trapeze Method on GPU and CPU to Solve a Physical Problem using CUDA

Youness Rtal, Abdelkader Hadjoudja

2021

Abstract

Graphical Processing Units (GPUs) are microprocessors attached to graphics cards and dedicated to displaying and manipulating graphic data. Currently, such graphics cards GPUs occupy all modern graphics cards. In a few years, these microprocessors have become potent, tools for parallel computing. Such processors are practical instruments that develop several fields like video and audio coding and decoding, resolution of a physical system to one or more unknowns. Their advantages are faster processing and lower power consumption than the power of the central processing unit (CPU). This paper will define and implement the Trapeze method to solve a problem in physics that computes the execution time of a vehicle moving at a speed v using the CUDA C/C++ parallel computing platform. This is a programming model from NVIDIA that increases computational performance by exploiting the power of GPUs. This type of calculation can be helpful, to control the speed of vehicles by radars with precision. The objective, of this study, is to compare the performance of the implementation of the Trapeze method on CPU and GPU processors and deduce the efficiency of using GPUs for parallel computing.

Download


Paper Citation


in Harvard Style

Rtal Y. and Hadjoudja A. (2021). Implementing the Trapeze Method on GPU and CPU to Solve a Physical Problem using CUDA. In Proceedings of the 2nd International Conference on Big Data, Modelling and Machine Learning - Volume 1: BML, ISBN 978-989-758-559-3, pages 45-49. DOI: 10.5220/0010728100003101


in Bibtex Style

@conference{bml21,
author={Youness Rtal and Abdelkader Hadjoudja},
title={Implementing the Trapeze Method on GPU and CPU to Solve a Physical Problem using CUDA},
booktitle={Proceedings of the 2nd International Conference on Big Data, Modelling and Machine Learning - Volume 1: BML,},
year={2021},
pages={45-49},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010728100003101},
isbn={978-989-758-559-3},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Big Data, Modelling and Machine Learning - Volume 1: BML,
TI - Implementing the Trapeze Method on GPU and CPU to Solve a Physical Problem using CUDA
SN - 978-989-758-559-3
AU - Rtal Y.
AU - Hadjoudja A.
PY - 2021
SP - 45
EP - 49
DO - 10.5220/0010728100003101