Effect And Analysis of Elastic Fidelity Computing On GPUs
Electrical Engineering & Computer Science Department, Northwestern University, Evanston, Illinois, USA
Northwestern University, Technical report, 2014
@techreport{patil2014effect,
title={Effect And Analysis of Elastic Fidelity Computing On GPUs},
author={Patil, Chetan Arvind and Tziantzioulis, Georgios and Hardavellas, Nikos},
year={2014}
}
The graphics processing unit (GPU) has become an integral part and plays a very vital role in high-end computing. Though GPU can easily reduce the execution time, it comes at the expense of power and energy consumption. There are various ways and approaches to reduce the power and energy consumption, Elastic Fidelity Computing (EFC) in one of them. EFC has been studied and used widely to reduce power and energy consumption in CPU. However, to the best of our knowledge it has not been tested so far on GPU. Therefore, there is a need to put forward ways to study the effects of EFC on GPU. In the last technical report, we introduce a way to study EFC on GPU by making use of standard GPU benchmarks and simulator. This report put forwards further progress done and important results achieved so far.
June 26, 2014 by hgpu
