Heterogeneous Parallel Computing for Data Encryption Application

Heterogeneous High Performance Computing (HPC) platforms consisting of multicore microprocessors and multicore GPU's are becoming increasingly popular these days. Many HPC and non-HPC applications are parallelized on these platforms. Data encryption is a common operation in a network based application programs with security. In order to keep pace with the high input data rate in such applications, real-time processing of data encryption is essential. In this paper, we propose a new approach to parallelize a data encryption algorithm, AES-CTR, on a heterogeneous HPC platform by extending the size of the block which is encrypted at one time across the unit block boundaries. The proposed approach leads to significant performance improvements.