An Adaptive and Integrated Low-Power Framework for Multicore Mobile Computing

Choi, Jongmoo; Jung, Bumjong; Choi, Yongjae; Son, Seiil

doi:10.1155/2017/9642958

Cited by 11 publications

(3 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At the software level, several works attempted to provide energy optimization methods at compilation and runtime. Choi et al [9] proposed an energy-aware framework by adapting Dynamic Power Management and DVFS policies with the required workload. This work achieved 22% to 79% power reduction while insignificantly affecting the workloads.…”

Section: Related Workmentioning

confidence: 99%

An in-Depth Evaluation of Frequency-Aware Scheduler for Improving User Experience on Mobile Devices

Tran

Carlier

Hagimont

2022

JCC

View full text Add to dashboard Cite

Mobile devices are more and more invading our daily life. Users of such devices expect to have a good experience, which is mainly linked with performance. However, higher performance also means a reduction in battery life, negatively contributing to the overall user experience. A common way to balance this performance-battery trade-off is to reduce CPU frequency when underload with Dynamic Voltage and Frequency Scaling. In our previous work, we introduced a Frequency-Aware Completely Fair Scheduler (called FA-CFS), which helps in reducing battery consumption and increasing the smoothness of mobile interface browsing. However, the current evaluation of FA-CFS model is only at quantitative results of power consumption rather than user experience on using their mobile device. In this paper, we perform an in-depth evaluation of the FA-CFS model, both quantitative results for system performance evaluation and qualitative results for user experience on mobile device usage. The experiments show that FA-CFS can reduce the rate of interface frame time peaks by up to 40% in terms of quantitative results and obtains a quantifiable impact on the quality of user experience with a quicker, more responsive interface.

show abstract

Section: Related Workmentioning

confidence: 99%

An in-Depth Evaluation of Frequency-Aware Scheduler for Improving User Experience on Mobile Devices

Tran

Carlier

Hagimont

2022

JCC

View full text Add to dashboard Cite

show abstract

“…In view of these benchmarking efforts, it is remarkable that thermal issues derived from running CNNs on edge devices have hardly been addressed in previous works. In [22], different algorithms were studied for managing both platform overheating and power consumption during extensive workloads on IoT devices. To this end, the authors applied well-known techniques such as dynamic power management and dynamic voltage and frequency scaling (DVFS).…”

Section: Related Workmentioning

confidence: 99%

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

Benoit-Cattin¹,

Velasco-Montero

Fernández-Berni

2020

Electronics

View full text Add to dashboard Cite

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.

show abstract

“…A security solution will not be accepted by consumers if it may cause tangible performance degradation [9][10][11][12]. Furthermore, the inclusion of multi-core CPUs and GPUs into small gadgets and low-power devices, such as smartphones, IoT, and embedded devices, can significantly help enhance device performance while ensuring data protection through encryption [13,14]. However, to take advantage of multi-core processors and improve performance in smart devices, a suitable parallel system needs to be built for that purpose [15,16].…”

Section: Introductionmentioning

confidence: 99%

Embedded Devices Security: Design and Implementation of a Light RDBMS Encryption Utilizing Multi-Core Processors

et al. 2023

View full text Add to dashboard Cite

The pervasive proliferation of embedded, mobile, and IoT devices continue to change our lifestyle dramatically. However, the huge increase in these devices has come with critical breaches to data resting inside them. Many types of such data are considered to be sensitive and confidential. Because the most sensitive data of such devices are resting in databases, focusing on encrypting SQLite databases will be more efficient than full disk encryption (FDE). While SQLite is a very popular, lightweight, and easy-to-use relational database suitable for embedded and mobile devices, its stored data suffers serious security risks. If an attacker can gain access to higher system privileges or find a way to access the database plain file, he can tamper with the database files and user-sensitive data, which breaches the security CIA triad of SQLite. To ensure data confidentiality in SQLite databases of embedded devices, we present a design and implementation of a parallel database encryption system, called SQLite-XTS. The developed system encrypts the database pages on-the-fly in a transparent manner without user intervention. Because performance is a critical issue, SQLite-XTS utilizes multi-core processors coming with most current mobile and embedded devices. The developed parallel SQLite-XTS was successfully implemented and integrated into a testbed device. To assess the performance and feasibility of this system, it was compared to three other SQLite implementations: plain SQLite, serial XTS SQLite, and SQLCipher-CBC. The results show that SQLite-XTS reduces the overhead of database encryption from 30.8% with serial implementation to 17.8% when SQLite-XTS is used. This provides the developed system with an efficiency of 73% compared with its serial counterpart. The results clarify that SQLite-XTS introduces significant performance improvements compared to other implementations. Experiments also show that the system has a very low impact on the memory of these resource-limited devices.

show abstract

An Adaptive and Integrated Low-Power Framework for Multicore Mobile Computing

Cited by 11 publications

References 20 publications

An in-Depth Evaluation of Frequency-Aware Scheduler for Improving User Experience on Mobile Devices

An in-Depth Evaluation of Frequency-Aware Scheduler for Improving User Experience on Mobile Devices

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

Embedded Devices Security: Design and Implementation of a Light RDBMS Encryption Utilizing Multi-Core Processors

Contact Info

Product

Resources

About