Comprehensive Analysis of the Uses of GPU and CUDA in Soft-Computing Techniques

Abstract

Now a day it is quite often seen that Soft-computing approaches are the cast off frequently in Computer and Information Technology. Fuzzy logic, Genetic Algorithms, and Artificial Neural Network are the fundamentals of Soft computing techniques. Several soft computing approaches such as deep neural network require a huge amount of computational power of the CPU to execute a problem and consume lots of time to execute. GPU (Graphics Processing Unit) provides an efficient way to use efficiently all the computational resources in the form of streaming processors. Further, CUDA (Computed Unified Device Architecture) provides a parallel framework specifically designed for GPU. CUDA and GPU are frequently used to implement different soft computing techniques. This article provides a brief introduction of GPU and CUDA and different soft computing techniques. Further, analytical review of the implementation of different soft computing techniques using GPU and CUDA are given.