gR: A GPU-based Router
Department of Information Science and Technology, Anna University, Chennai, India
IEEE International Conference on High Performance Computing (HiPC), 2012
@article{sundaresan2012gr,
title={gR: A GPU-based Router},
author={Sundaresan, P. and Venkateshprasad, S. and Muralitharan, Y. and Parthasarathi, R.},
year={2012}
}
With the growing internet traffic and complexity of packet processing task, the throughput of routers is affected. Also modern routers need to provide additional services like security, QOS which further adds to the complexity. These issues can be addressed with the massive parallel computing capability of graphic processors. In this paper, we offload two of the most computationally intensive tasks namely lookup and two-key IPSec to CUDA-enabled GPU. The execution time is reduced by exploiting the inherent parallelism in packet processing tasks. Finally, we have analyzed the performance variations and have found that modern graphic cards can be effectively used to speed up router operations.
December 4, 2012 by hgpu
