TY - JOUR
T1 - Dynamic Resource Management of Heterogeneous Mobile Platforms via Imitation Learning
AU - Mandal, Sumit K.
AU - Bhat, Ganapati
AU - Patil, Chetan Arvind
AU - Doppa, Janardhan Rao
AU - Pande, Partha Pratim
AU - Ogras, Umit Y.
N1 - Funding Information:
Manuscript received February 21, 2019; revised May 21, 2019; accepted June 14, 2019. Date of publication July 23, 2019; date of current version November 22, 2019. This work was supported in part by the National Science Foundation under Grant CNS-1526562, Grant OAC-1910213, and Grant IIS-1845922, in part by the Semiconductor Research Corporation (SRC) under Grant 2721.001, and in part by the U.S. Army Research Office under Grant W911NF-17-1-0485 and Grant W911NF-19-1-0162. (Corresponding author: Sumit K. Mandal.) S. K. Mandal, G. Bhat, and U. Y. Ogras are with the School of Electrical, Computer and Energy Engineering, Arizona State University, Tempe, AZ 85281 USA (e-mail: skmandal@asu.edu; gmbhat@asu.edu; umit@asu.edu).
Publisher Copyright:
© 1993-2012 IEEE.
PY - 2019/12
Y1 - 2019/12
N2 - The complexity of heterogeneous mobile platforms is growing at a rate faster than our ability to manage them optimally at runtime. For example, state-of-the-art systems-on-chip (SoCs) enable controlling the type (Big/Little), number, and frequency of active cores. Managing these platforms becomes challenging with the increase in the type, number, and supported frequency levels of the cores. However, existing solutions used in mobile platforms still rely on simple heuristics based on the utilization of cores. This paper presents a novel and practical imitation learning (IL) framework for dynamically controlling the type (Big/Little), number, and the frequencies of active cores in heterogeneous mobile processors. We present efficient approaches for constructing an Oracle policy to optimize different objective functions, such as energy and performance per Watt (PPW). The Oracle policies enable us to design low-overhead power management policies that achieve near-optimal performance matching the Oracle. Experiments on a commercial platform with 19 benchmarks show on an average 101% PPW improvement compared to the default interactive governor.
AB - The complexity of heterogeneous mobile platforms is growing at a rate faster than our ability to manage them optimally at runtime. For example, state-of-the-art systems-on-chip (SoCs) enable controlling the type (Big/Little), number, and frequency of active cores. Managing these platforms becomes challenging with the increase in the type, number, and supported frequency levels of the cores. However, existing solutions used in mobile platforms still rely on simple heuristics based on the utilization of cores. This paper presents a novel and practical imitation learning (IL) framework for dynamically controlling the type (Big/Little), number, and the frequencies of active cores in heterogeneous mobile processors. We present efficient approaches for constructing an Oracle policy to optimize different objective functions, such as energy and performance per Watt (PPW). The Oracle policies enable us to design low-overhead power management policies that achieve near-optimal performance matching the Oracle. Experiments on a commercial platform with 19 benchmarks show on an average 101% PPW improvement compared to the default interactive governor.
KW - Heterogeneous computing
KW - imitation learning (IL)
KW - multi-processor systems-on-chip (SoCs)
KW - multicore architectures
UR - http://www.scopus.com/inward/record.url?scp=85073074881&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85073074881&partnerID=8YFLogxK
U2 - 10.1109/TVLSI.2019.2926106
DO - 10.1109/TVLSI.2019.2926106
M3 - Article
AN - SCOPUS:85073074881
SN - 1063-8210
VL - 27
SP - 2842
EP - 2854
JO - IEEE Transactions on Very Large Scale Integration (VLSI) Systems
JF - IEEE Transactions on Very Large Scale Integration (VLSI) Systems
IS - 12
M1 - 8770273
ER -