Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces a methodology for accurate CPU power estimation on heterogeneous ARM-based mobile devices in Federated Learning (FL) settings, using a rail-to-cluster mapping technique to retrieve cluster-level supply voltage for analytical CMOS-based power modeling. The methodology is validated on Android devices, demonstrating that the analytical model achieves <10% error in power prediction, significantly outperforming approximate models (up to 959% error). Integrating the analytical model into AnycostFL resulted in 1.4x energy savings while maintaining 80% model accuracy, highlighting the importance of accurate power modeling in energy-aware FL.
In energy-aware Federated Learning, inaccurate power models can lead to nearly 10x overestimation of energy consumption and suboptimal resource allocation.
Estimating CPU power on heterogeneous ARM-based commodity devices is challenging due to limited access to CPU's voltage domains. As a result, state-of-the-art energy-aware Federated Learning (FL) frameworks typically rely on simplified approximate power models to estimate computation energy, rather than the more accurate analytical CMOS-based model. To bridge this gap, we propose a reproducible CPU power estimation methodology combined with a rail-to-cluster mapping technique to retrieve cluster-level supply voltage. We evaluate our approach on two commodity Android devices and show that the analytical model predicts CPU power with errors below 10%, whereas the approximate model incurs errors of up to 959%. Using AnycostFL, a state-of-the-art energy-aware FL framework, we show that the analytical model achieves the same 80% model accuracy while consuming 1.4x less energy than the approximate model. These results highlight that approximate models can severely misestimate computation energy and lead to suboptimal decisions. This work facilitates the use of analytical CPU power models on heterogeneous multi-cluster ARM-based mobile SoCs without additional hardware support or external power measurement tools.