Runtime Support for Adaptive Power Capping on Heterogeneous SoCs
Published in Proceedings of the International Conference on Embedded Computer Systems: Architecture, Modeling and Simulation (SAMOS XVI), 71–78, 2017
This paper addresses the challenge of managing power consumption in heterogeneous System-on-Chip (SoC) architectures by introducing a runtime system that enables adaptive power capping. The proposed system dynamically adjusts the power budget allocated to different components of the SoC, balancing performance and power consumption to meet system-level constraints.
Key contributions include:
Adaptive Power Capping: Development of a runtime system that monitors and adjusts the power consumption of heterogeneous SoC components, ensuring compliance with power budgets while maintaining performance.
Performance-Power Trade-offs: Analysis of the trade-offs between performance and power consumption, providing insights into how adaptive power capping can be leveraged to optimize system behavior.
Implementation and Evaluation: Implementation of the proposed system on a heterogeneous SoC platform and evaluation of its effectiveness in real-world scenarios, demonstrating its capability to adapt to varying workloads and power constraints.
Recommended citation: Wu, Y., Nikolopoulos, D. S., & Woods, R. (2017). Runtime Support for Adaptive Power Capping on Heterogeneous SoCs. In *Proceedings of the International Conference on Embedded Computer Systems: Architecture, Modeling and Simulation (SAMOS XVI)* (pp. 71–78). IEEE. https://doi.org/10.1109/SAMOS.2017.7818333
Download Paper | Download Slides | Download Bibtex