Abstract
This paper presents a platform-level power management framework for mobile platforms. The proposed framework minimizes the overall platform energy while meeting system level performance and power budget constraints. To this end, we construct analytical performance and power models using dynamic information collected via performance monitoring counters. Using these models, we design two different closed loop controllers to ensure that both the performance and the power targets are achieved and maintained in the presence of dynamic workload variations. Experimental evaluations performed on an Android platform show up to 8% energy savings at the platform level and up to 15% CPU energy savings.
| Original language | English (US) |
|---|---|
| Title of host publication | International System on Chip Conference |
| Publisher | IEEE Computer Society |
| Pages | 146-151 |
| Number of pages | 6 |
| ISBN (Print) | 9781479933785 |
| DOIs | |
| State | Published - Nov 5 2014 |
| Event | 27th IEEE International System on Chip Conference, SOCC 2014 - Las Vegas, United States Duration: Sep 2 2014 → Sep 5 2014 |
Other
| Other | 27th IEEE International System on Chip Conference, SOCC 2014 |
|---|---|
| Country | United States |
| City | Las Vegas |
| Period | 9/2/14 → 9/5/14 |
Fingerprint
ASJC Scopus subject areas
- Hardware and Architecture
- Control and Systems Engineering
- Electrical and Electronic Engineering
Cite this
Towards platform level power management in mobile systems. / Kadjo, David; Ogras, Umit; Ayoub, Raid; Kishinevsky, Michael; Gratz, Paul.
International System on Chip Conference. IEEE Computer Society, 2014. p. 146-151 6948916.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Towards platform level power management in mobile systems
AU - Kadjo, David
AU - Ogras, Umit
AU - Ayoub, Raid
AU - Kishinevsky, Michael
AU - Gratz, Paul
PY - 2014/11/5
Y1 - 2014/11/5
N2 - This paper presents a platform-level power management framework for mobile platforms. The proposed framework minimizes the overall platform energy while meeting system level performance and power budget constraints. To this end, we construct analytical performance and power models using dynamic information collected via performance monitoring counters. Using these models, we design two different closed loop controllers to ensure that both the performance and the power targets are achieved and maintained in the presence of dynamic workload variations. Experimental evaluations performed on an Android platform show up to 8% energy savings at the platform level and up to 15% CPU energy savings.
AB - This paper presents a platform-level power management framework for mobile platforms. The proposed framework minimizes the overall platform energy while meeting system level performance and power budget constraints. To this end, we construct analytical performance and power models using dynamic information collected via performance monitoring counters. Using these models, we design two different closed loop controllers to ensure that both the performance and the power targets are achieved and maintained in the presence of dynamic workload variations. Experimental evaluations performed on an Android platform show up to 8% energy savings at the platform level and up to 15% CPU energy savings.
UR - http://www.scopus.com/inward/record.url?scp=84911891948&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84911891948&partnerID=8YFLogxK
U2 - 10.1109/SOCC.2014.6948916
DO - 10.1109/SOCC.2014.6948916
M3 - Conference contribution
SN - 9781479933785
SP - 146
EP - 151
BT - International System on Chip Conference
PB - IEEE Computer Society
ER -