A Survey on AES (Advanced Encryption Standard) and RSA Encryption-Decryption in CUDA

Abstract

CPUs and GPUs have fundamentally different design philosophies, but combining their characteristics could avail better performance and throughput. In this paper, we will study the technologies of GPU parallel computing and its optimized design for cryptography.The advent of the Compute Unified Device Architecture (CUDA) from NVIDIA Technology has shifted Graphics Processing Units (GPUs) from primarily graphics enabling devices to general purpose stream processing devices. Cryptography is the study of techniques focused on security. Two algorithms are selected for investigation AES and RSA in CUDA. The designs of most cryptographic algorithms are such that they can benefit considerably from parallel computing, which consumer GPUs can provide inexpensively and economically. Parallelization of these security algorithms in order to distribute the complex computational part among the various cores available with the processors today, will achieve higher performance and also be more energy efficient. Hence combination of both algorithms will help to provide better security and efficiency.