TY - GEN
T1 - A measurement study of GPU DVFS on energy conservation
AU - Mei, Xinxin
AU - Yung, Ling Sing
AU - Zhao, Kaiyong
AU - CHU, Xiaowen
PY - 2013
Y1 - 2013
N2 - Nowadays, GPUs are widely used to accelerate many high performance computing applications. Energy conservation of such computing systems has become an important research topic. Dynamic voltage/frequency scaling (DVFS) is proved to be an appealing method for saving energy for traditional computing centers. However, there is still a lack of firsthand study on the effectiveness of GPU DVFS. This paper presents a thorough measurement study that aims to explore how GPU DVFS affects the system energy consumption. We conduct experiments on a real GPU platform with 37 benchmark applications. Our results show that GPU voltage/frequency scaling is an effective approach to conserving energy. For example, by scaling down the GPU core voltage and frequency, we have achieved an average of 19.28% energy reduction compared with the default setting, while giving up no more than 4% of performance. For all tested GPU applications, core voltage scaling is significantly effective to reduce system energy consumption. Meanwhile the effects of scaling core frequency and memory frequency depend on the characteristics of GPU applications.
AB - Nowadays, GPUs are widely used to accelerate many high performance computing applications. Energy conservation of such computing systems has become an important research topic. Dynamic voltage/frequency scaling (DVFS) is proved to be an appealing method for saving energy for traditional computing centers. However, there is still a lack of firsthand study on the effectiveness of GPU DVFS. This paper presents a thorough measurement study that aims to explore how GPU DVFS affects the system energy consumption. We conduct experiments on a real GPU platform with 37 benchmark applications. Our results show that GPU voltage/frequency scaling is an effective approach to conserving energy. For example, by scaling down the GPU core voltage and frequency, we have achieved an average of 19.28% energy reduction compared with the default setting, while giving up no more than 4% of performance. For all tested GPU applications, core voltage scaling is significantly effective to reduce system energy consumption. Meanwhile the effects of scaling core frequency and memory frequency depend on the characteristics of GPU applications.
KW - energy conservation
KW - GPU
KW - voltage/frequency scaling
UR - http://www.scopus.com/inward/record.url?scp=84889632170&partnerID=8YFLogxK
U2 - 10.1145/2525526.2525852
DO - 10.1145/2525526.2525852
M3 - Conference proceeding
AN - SCOPUS:84889632170
SN - 9781450324588
T3 - Proceedings of the Workshop on Power-Aware Computing and Systems, HotPower 2013
BT - Proceedings of the Workshop on Power-Aware Computing and Systems, HotPower 2013
T2 - Workshop on Power-Aware Computing and Systems, HotPower 2013
Y2 - 3 November 2013 through 6 November 2013
ER -
