Abstract

GPUs (Graphics Processing Units) are microprocessors attached to graphics cards dedicated to the display and manipulation of graphics data. In a few years, these microprocessors (GPUs) have occupied all modern graphics cards and become very important tools for massively parallel computing. These processors are practical tools for the development of several areas such as image processing, video and audio coding and decoding, solving a physical system with one or more unknowns... Their advantages: faster processing and lower energy consumption than the power of the central processing unit (CPU). In this paper, we will define and implement the Lagrange interpolation method on GPU and CPU to compute the density of a metal at different Ti temperatures using the CUDA C parallel programming model from NVIDIA which is used to increase the computational performance by exploiting the power of the GPU. Our goal is to compare the performance of the Lagrange interpolation method implementation on CPU and GPU processors and to infer the efficiency of using GPUs for parallel computing.