Predictive Dynamic Thermal and Power Management for Heterogeneous Mobile Platforms

Singla, Gaurav; Kaur, Gurinderjit; Unver, Ali K.; Ogras, Ümit Y.

doi:10.7873/date.2015.1036

Cited by 83 publications

(63 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Predictive approaches develop computationally efficient thermal and power models [17,21]. These models are used at runtime to estimate the short-term behavior of the power-temperature dynamics and take actions if violations are predicted [21,22,23]. These approaches are effective in predicting the short-term behavior of the power-temperature dynamics; however, the prediction error increases when long-term predictions are made [23].…”

Section: Related Workmentioning

confidence: 99%

“…These models are used at runtime to estimate the short-term behavior of the power-temperature dynamics and take actions if violations are predicted [21,22,23]. These approaches are effective in predicting the short-term behavior of the power-temperature dynamics; however, the prediction error increases when long-term predictions are made [23]. A number of studies have also used control-theoretic approaches to regulate the temperature of the system [24,25,26,27].…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Analysis and Control of Power–Temperature Dynamics in Heterogeneous Multiprocessors

Bhat

Gümüşsoy

Ogras

2021

IEEE Trans. Contr. Syst. Technol.

Self Cite

View full text Add to dashboard Cite

Virtually all electronic systems try to optimize a fundamental trade-off between higher performance and lower power consumption. The latter becomes critical in mobile computing systems, such as smartphones, which rely on passive cooling. Otherwise, the heat concentrated in a small area drives both the junction and skin temperatures up. High junction temperatures degrade the reliability, while skin temperature deteriorates the user experience. Therefore, there is a strong need for a formal analysis of power consumption-temperature dynamics and predictive thermal management algorithms. This paper presents a theoretical power-temperature analysis of multiprocessor systems, which are modeled as multi-input multi-output dynamic systems. We analyze the conditions under which the system converges to a stable steady-state temperature. Then, we use these models to design a control algorithm that manages the temperature of the system without affecting the performance of the application. Experiments on the Odroid-XU3 board show that the control algorithm is able to regulate the temperature with a minimal loss in performance when compared to the default thermal governors.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Analysis and Control of Power–Temperature Dynamics in Heterogeneous Multiprocessors

Bhat

Gümüşsoy

Ogras

2021

IEEE Trans. Contr. Syst. Technol.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Although power management approaches could, to some extent, alleviate the thermal hot spots across the chip, increasing power density of MPSoCs made bare power management insufficient to deal with hot spots and led authors to propose thermal management policies at both design [23]- [25] and run time [26]- [31]. In particular, [23] and [24] propose optimal solutions for task scheduling and processor speed, respectively, [25] maximizes the performance of a periodic application, and [26] presents thermal balancing policy.…”

Section: B Power and Thermal Managementmentioning

confidence: 99%

“…The thermal impacts of the adjacent cores on the thermal profile is considered in [30]. Authors in [31] propose a dynamic thermal and power management using temperature prediction methodology. All these works, however, fail to consider thermal stress as a new dominant factor in modern MPSoCs lifetime reliability [6].…”

Section: B Power and Thermal Managementmentioning

confidence: 99%

TheSPoT: Thermal Stress-Aware Power and Temperature Management for Multiprocessor Systems-on-Chip

Iranfar

Kamal

Afzali-Kusha

et al. 2018

IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.

View full text Add to dashboard Cite

Abstract-Thermal stress including temperature gradients in time and space, as well as thermal cycling, influences lifetime reliability and performance of modern Multiprocessor Systemson-Chip (MPSoCs). Conventional power and temperature management techniques considering the peak temperature/power consumption do not provide a comprehensive solution to avoid high spatial and temporal thermal variations. This work presents TheSPoT, a novel multi-level thermal stress-aware power and thermal management approach for MPSoCs. At the top level, core consolidation and deconsolidation is performed based on peak temperature, thermal stress, and power consumption constraints. These constraints are also used at the next level, where operating frequencies are determined. At this level we obtain optimal core frequencies by solving a convex optimization problem. However, thereafter, to reduce the runtime overhead in large MPSoCs, we alternatively propose to use a fast heuristic algorithm. The efficacy of the proposed approaches in reducing the thermal cycles and temporal/spatial temperature gradients is evaluated by comparing the results with the state-of-the-art methods. The evaluation performed on 4-core, 8-core, and 16-core MPSoCs, using PARSEC benchmarks, reveals a considerable reduction in thermal stress. For the 8-core MPSoC case study, on average, for the proposed heuristic(optimal) approach, the mean time to failure improved by 47(35) % compared to the state-of-the-art techniques with only 6(4) % performance degradation. Also, our simulations show that TheSPoT is more efficient in thermal stress reduction when more heterogeneous workloads are used.

show abstract

“…ARM big.LITTLE TM technology [4]). It has been shown that run-time management software (or the run-time manager) can make significant energy improvements by controlling these power-saving techniques to smartly manage the energy-performance trade-off while taking external factors into account [5], [6]. However, run-time knowledge of the power consumption of each CPU core in the system is paramount in finding the optimum power/performance trade-off.…”

Section: Introductionmentioning

confidence: 99%

Accurate and Stable Run-Time Power Modeling for Mobile and Embedded CPUs

Walker

Diestelhorst

Hansson

et al. 2017

IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.

105

View full text Add to dashboard Cite

Abstract-Modern mobile and embedded devices are required to be increasingly energy-efficient while running more sophisticated tasks, causing the CPU design to become more complex and employ more energy-saving techniques. This has created a greater need for fast and accurate power estimation frameworks for both run-time CPU energy management and design-space exploration. We present a statistically rigorous and novel methodology for building accurate run-time power models using Performance Monitoring Counters (PMCs) for mobile and embedded devices, and demonstrate how our models make more efficient use of limited training data and better adapt to unseen scenarios by uniquely considering stability. Our robust model formulation reduces multicollinearity, allows separation of static and dynamic power, and allows a 100× reduction in experiment time while sacrificing only 0.6% accuracy. We present a statistically detailed evaluation of our model, highlighting and addressing the problem of heteroscedasticity in power modeling. We present software implementing our methodology and build power models for ARM Cortex-A7 and Cortex-A15 CPUs, with 3.8% and 2.8% average error, respectively. We model the behavior of the nonideal CPU voltage regulator under dynamic CPU activity to improve modeling accuracy by up to 5.5% in situations where the voltage cannot be measured. To address the lack of research utilizing PMC data from real mobile devices, we also present our data acquisition method and experimental platform software. We support this work with online resources including software tools, documentation, raw data and further results.

show abstract

Predictive Dynamic Thermal and Power Management for Heterogeneous Mobile Platforms

Cited by 83 publications

References 33 publications

Analysis and Control of Power–Temperature Dynamics in Heterogeneous Multiprocessors

Analysis and Control of Power–Temperature Dynamics in Heterogeneous Multiprocessors

TheSPoT: Thermal Stress-Aware Power and Temperature Management for Multiprocessor Systems-on-Chip

Accurate and Stable Run-Time Power Modeling for Mobile and Embedded CPUs

Contact Info

Product

Resources

About