Impact of Thermal Throttling on Long-Term Visual Inference in a CPU-Based Edge Device

Abstract

Many application scenarios of edge visual inference, e.g., robotics or environmental monitoring, eventually require long periods of continuous operation. In such periods, the processor temperature plays a critical role to keep a prescribed frame rate. Particularly, the heavy computational load of convolutional neural networks (CNNs) may lead to thermal throttling and hence performance degradation in few seconds. In this paper, we report and analyze the long-term performance of 80 different cases resulting from running five CNN models on four software frameworks and two operating systems without and with active cooling. This comprehensive study was conducted on a low-cost edge platform, namely Raspberry Pi 4B (RPi4B), under stable indoor conditions. The results show that hysteresis-based active cooling prevented thermal throttling in all cases, thereby improving the throughput up to approximately 90% versus no cooling. Interestingly, the range of fan usage during active cooling varied from 33% to 65%. Given the impact of the fan on the power consumption of the system as a whole, these results stress the importance of a suitable selection of CNN model and software components. To assess the performance in outdoor applications, we integrated an external temperature sensor with the RPi4B and conducted a set of experiments with no active cooling in a wide interval of ambient temperature, ranging from 22 °C to 36 °C. Variations up to 27.7% were measured with respect to the maximum throughput achieved in that interval. This demonstrates that ambient temperature is a critical parameter in case active cooling cannot be applied.

Design and Evaluation of a Heterogeneous Lightweight Blockchain-Based Marketplace

“…With respect to CPU usage in these compact devices, it is necessary to consider if the device would suffer from high-temperature issues. These devices are known to work at high temperatures, and high CPU usage would cause temporary or permanent undesirable effects, such as throttling [ 61 ]. Due to that, CPU temperatures were monitored every half second in order to control any possible risk if the device reaches high CPU temperature.…”

Optical calibration and first light for the deformable mirror demonstration mission CubeSat (DeMi)

“…During development, the DeMi team found that using Raspberry Pi's as the payload computers provided an accessible computing interface for software development. However, the Raspberry Pi 3 compute modules used in the payload exhibit strong heating under vacuum conditions, which causes throttling of the communications interfaces 41,42 and needed to be mitigated with software changes and regular power cycles. This throttling effect has been observed during on-orbit operations as well.…”

Metareasoning Approaches for Thermal Management During Image Processing

“…Benoit-Cattin et al [11] showed that dynamic active cooling can increase the efficiency of image processing on a Raspberry Pi. Their approach, which used an external fan that was controlled by the Raspberry Pi 4B, increased the image processing throughput.…”