It is well known that the cost of executing the sequential portion of a program will limit and sometimes even eclipse the gains brought by processing in parallel the rest of the program. This means that serious consideration should be brought to bear on accelerating the execution of this unavoidable sequential part. Such acceleration can be done by boosting the operating frequency in a symmetric multicore processor. In this paper, we derive a performance and power model to describe the implications of this approach. From our model, we show that the ratio of performance over energy during the sequential part improves with an increase in the number of cores. In addition, we demonstrate how to determine with the proposed model the optimal frequency boosting ratio which maximizes energy efficiency.
Bibliographical noteFunding Information:
This work was supported in part by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2010-0013202) and by the National Science Foundation, under award CCF-1439165. Any opinions, findings, and conclusions expressed in this material are those of the authors and do not necessarily reflect the views of the sponsors. W.W. Ro is the corresponding author.
All Science Journal Classification (ASJC) codes
- Hardware and Architecture