Thermal Aware Research Articles

Minimum-energy virtual machine placement using embedded sensors and machine learning

Cloud data centers (DCs) consume large amounts of energy and contribute significantly to environmental concerns. Furthermore, with the advent of 5G and B5G networks, increasingly software-oriented and becoming highly dependent on cloud computing, it becomes imperative to optimize their energy consumption. Thus, in this study, we present a virtual machine placement algorithm that minimizes the energy consumption of a cluster of server machines. Our solution is embodied through the use of sensors embedded inside physical server machines, enabling the introduction of new features for sensitive thermal awareness and proactive hot spot avoidance. Leveraging this significantly enhanced feature space, we implement data-driven predictive machine learning models along with a heuristic placement algorithm (CPP), enabling proactive VM placements that are both energy-aware and thermal-aware. Indeed, experiments carried out on a cluster of physical server machines demonstrate high performance by both the ML models and the placement algorithm (CPP). Compared with the best baseline algorithm, our solution reduced power consumption and temperature by 7% and 2%, respectively, while avoiding hot spots and maintaining efficient load distribution, thereby reducing the overhead of physical machines by 28%.

The thermoception task: a thermal imaging-based procedure for measuring awareness of changes in peripheral body temperature.

Although thermal body signals provide crucial information about the state of an organism and changes in body temperature may be a sign of affective states (e.g., stress, pain, sexual arousal), research on thermal awareness is limited. Here we developed a task measuring awareness of changes in peripheral body temperature (thermal interoception) and compared it to the classical heartbeat counting task (cardiac interoception). With an infrared light bulb we delivered stimuli of different temperature intensities to the right hand of 31 healthy participants. Thermal interoceptive accuracy, i.e., the difference between participants' real and perceived change in hand temperature, showed good interindividual variability. We found that thermal interoception did not correlate with (and was generally higher than) cardiac interoception, suggesting that different interceptive channels provide separate contributions to awareness of bodily states. Moreover, the results hint at the great salience of thermal signals and the need for thermoregulation in day-to-day life. Finally, thermal interoceptive accuracy was associated with self-reported awareness of body temperature changes and with the ability to regulate distress by focusing on body sensations. Our task has the potential to significantly increase current knowledge about the role of interoception in cognition and behavior, particularly in social and emotional contexts.NEW & NOTEWORTHY We developed a novel task measuring awareness of changes in peripheral body temperature (i.e., thermal interoception). To avoid tactile confounds present in existing thermoceptive tasks, we used an infrared light bulb to deliver stimuli of different temperature intensities to the hand of participants and asked them to judge the perceived change in their hand temperature. Performance in the task showed good interindividual variability, did not correlate with cardiac interoceptive tasks, and was associated with self-reported thermosensitivity.

Towards Thermal-Aware Workload Distribution in Cloud Data Centers Based on Failure Models

Increasing workload conditions lead to a significant surge in power consumption and computing node failures in data centers. The existing workload distribution strategies focused on either thermal awareness or failure mitigation, overlooking the impact of node failures on the energy efficiency of cloud data centers. To address this issue, a new holistic model is built to characterize the impacts of workloads, computing and cooling costs, heat recirculation, and node failure on the energy efficiency of cloud data centers. Leveraging such a holistic model, we propose a novel thermal-aware workload distribution strategy called <i>HGSA</i> that takes node failure into accountand can improve the energy efficiency of cloud data centers. Our empirical findings confirm that (i) faulty nodes lead to a large rise in power consumption, and (ii) failure locations play a vital role in the power consumption of data centers. Experimental results unveil that HGSA is adroit at making near-optimal decisions in workload distribution strategies. In particular, HGSA cuts down the minimum inlet temperature by 5.2 <inline-formula><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> -15 <inline-formula><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> , improves the maximum air temperature of a Computer Room Air Conditioner (CRAC) model by 4.2 <inline-formula><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> -26.5 <inline-formula><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> , lowers the cooling cost by 15.4 <inline-formula><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> -50 <inline-formula><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> compared to the existing solutions. Furthermore, HGSA cuts back the total power consumption by 0.65 <inline-formula><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> -78 <inline-formula><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> .

The Thermal Responses between Young Adults and Preschool Children in a Radiant Floor Heating Environment

The thermal comfort of preschool children was assumed to be similar to that of young adults, which may cause inaccuracy. This study tested and analyzed the thermal response characteristics of young adults and preschool children (4–6 years old) and the differences in thermal sensation and thermal physiology between the two groups of participants in a room with a radiant floor heating system using the difference analysis methods (the paired data t-test, the Mann–Whitney U test and the Kruskal–Wallis H test). Participants were divided into two groups, young adults and preschoolers, and were sat in each condition while wearing winter clothing with a thermal resistance of 1.02 clo. The results showed that when the indoor temperature changed, there was a significant difference in the local skin temperature of the calf between the two groups of participants (p &lt; 0.05). Preschool children adapt to the thermal environment better than adults, and the difference in metabolic rate is one of the influencing factors. The overall thermal sensation with mean skin temperature of the different populations was linearly correlated; correlation coefficients were 0.944 and 0.932, respectively. The overall thermal sensation of the participants was linear with respect to the indoor operative temperature. Preschool children have a higher thermal sensitivity to temperature change than young adults under low-temperature radiant floor heating systems, indicating that children have different thermal awareness from adults. There were significant differences in preschoolers’ subjective assessments of thermal sensation when the predicted mean vote (PMV) model was used as the evaluation standard; the difference ranged from 0.77 to 2.33. Thus, the PMV-predicted percentage dissatisfied (PPD) model is not suitable for preschool children.

Server temperature prediction using deep neural networks to assist thermal-aware scheduling

Thermal-Aware (TA) scheduling is a primitive thermal-management tool to avoid hotspots and attain a better thermal profile inside data centers. However, TA scheduling depends on the accuracy of server temperature calculation for making reliable scheduling decisions. Existing TA simulation frameworks rely on the CRAC, RC, and thermodynamics model to calculate the temperature of a server. However, these models are not very efficient computationally while calculating the temperature of computing nodes. Hence, there is a need for an efficient and lightweight temperature prediction model to fill this gap. Moreover, existing TA allocation and migration schemes neglect the workload heterogeneity regarding execution time (length) of user tasks in batch workloads. Ignoring task heterogeneity may lead to higher ambient temperature on its neighboring servers—assigning a hotter job with a longer duration to a server with higher ambient effect results in significantly higher cooling expenses.Our contribution in this paper is twofold: (1) firstly, we design and train a deep neural network to predict the temperature of servers; our proposed DNN model achieves 96.11% prediction accuracy, and (2) secondly, we propose a TA algorithm for job allocation and migration. Specifically, we consider the length and thermal profile of user tasks and servers during allocation and migration. We compare the proposed strategy against existing TA scheduling designs such as TA Scheduling Algorithm (TASA) and TA Control Strategy (TACS) using simulations. Results demonstrate that the proposed TA approach significantly reduces the overall energy consumption by up to 12.03% and 8.28% compared to the TASA and TACS, respectively.

Thermal Aware System-Wide Reliability Optimization for Automotive Distributed Computing Applications

As the automotive industry is shifting the paradigm towards autonomous driving, safety guarantee has become a paramount consideration. Temperature plays a key role in the system-wide reliability of the electronic control systems (ECS) used in the automotive. A vehicle is usually subjected to harsh temperature conditions from its operating environment. The increasing power density of IC chips in the ECS further exacerbates the operating temperature and thermal gradient condition on the chip, thereby significantly impacting the vehicle’s reliability. In this paper, we study how to map a periodic distributed automotive application on a heterogeneous multiple-core processing architecture with temperature and system-level reliability issues in check. We first present a mathematical programming model to bound the peak operating temperature for the ECS. Then we propose a more sophisticated approach based on the genetic algorithm to effectively bound the peak temperature and optimize the system-wide reliability of the ECS by maximizing its mean-time-to-failure (MTTF). To this end, we present an algorithm to guarantee the peak temperature for periodic applications with variable execution times to ensure our approach’s effectiveness. We also present several computationally efficient techniques for system-wide MTTF computation, which show several-order-of-magnitude speed-up over the state-of-the-art method when tested using synthetic cases and practical benchmarks.

Frequency-scaled thermal-aware test scheduling for 3D ICs using machine learning based temperature estimation

Three-dimensional Integrated Circuits (3D IC) are being looked at as an alternative in overcoming some critical challenges plaguing conventional 2D ICs. However, it faces some new challenges — some of them being thermal management and a complicated testing process. This work presents a frequency-scaled thermal-aware test scheduling algorithm for 3D ICs. The thermal awareness is provided by a proposed machine learning based predictive thermal estimation model for 3D ICs. The proposed test scheduling algorithm includes test clock frequency as an important parameter and scales the frequency to achieve lower test time. Experimental results for both the predictive thermal model and scheduling algorithm look promising, with improvements over existing works achieved in most of the cases. The thermal model estimates the temperature of a 3D IC with a high degree of accuracy but at a fraction of time of commonly used 3D IC thermal models, while the scheduling algorithm has been observed to have improved test time of different benchmark circuits up to around 60%. Experimental results show — that test time in terms of clock cycles is inconsequential because schedules with higher clock count may require less time due to frequency scaling.

Thermal Aware Routing Protocols for Wireless Body Area Networks: Review and Open Research Issues

Wireless Body Area Network (WBAN) is a promising technology that improves life quality and enhances medical healthcare services. The activities and communications of sensors lead to a temperature rise, which can be fatal for the human body. The heating problem is addressed by the thermal-aware strategy, which has great importance. This paper reviews sixteen thermal-aware routing protocols proposed for WBAN; it presents routing in WBAN and its associated challenges. It explains the principles and pros and cons of each protocol and compares the protocols studied using several criteria. Finally, the paper points out the open research issues and challenges. This review intends to provide a comprehensive, up-to-date reference for everyone interested in the thermal-aware communication protocols for WBANs.

Research advances on AI-powered thermal management for data centers

Given the complex nature of data centers' thermal management, which costs too many resources, processing time, and energy consumption, thermal awareness and thermal management powered by artificial intelligence (AI) are the targeted study. In addition to a few research on AI techniques and models, other strategies have also been introduced in recent years. Data center models, including cooling, thermal, power, and workload models, and their relationship are factors that need to be understood in the optimal thermal management system. Simulation approaches have been proposed to help validate new models or methods used for scheduling and consolidating processes and virtual machines (VMs), hotspot identification, thermal state estimation, and power usage change. AI-powered thermal optimization leads to improved process scheduling and consolidation of VMs and eliminates the hotspot from happening. At present, research on AI-powered thermal control is still in its infancy. This paper concludes with four issues in thermal management, which will be the scope of further research.

Thermal awareness to enhance data center energy efficiency

Data centers aim at provisioning on-demand processing, storage and networking capabilities in a reliable and scalable way. In this context, proper maintenance of IT equipment within data center (DC) premises is crucial as it ensures prolonged lifetime of servers and uninterrupted availability of resources. This paper focuses on the analysis how the exact temperatures in a real big DC cluster of 10 368 cores correspond to thermal guidelines for IT devices. Available readings of servers exhaust air temperature, inlet air temperature, as well as CPUs temperature have been analyzed to uncover general statistical properties which are subsequently aggregated into several descriptive metrics that reveal global and local trends. The ultimate outcome of this research is the improvement of DC thermal management for sustainable operations via a set of recommendations.

Optimization and Controlled Systems: A Case Study on Thermal Aware Workload Dispatching

Although successfully employed on many industrial problems, Combinatorial Optimization still has limited applicability on several real-world domains, often due to modeling difficulties. This is typically the case for systems under the control of an on-line policy: even when the policy itself is well known, capturing its effect on the system in a declarative model is often impossible by conventional means. Such a difficulty is at the root of the classical, sharp separation between off- line and on-line approaches. In this paper, we investigate a general method to model controlled systems, based on the integration of Machine Learning and Constraint Programming (CP). Specifically, we use an Artificial Neural Network (ANN) to learn the behavior of a controlled system (a multicore CPU with thermal con- trollers) and plug it into a CP model by means of Neuron Constraints. The method obtains significantly better results compared to an approach with no ANN guidance. Neuron Constraints were first introduced in [Bartolini et al., 2011b] as a mean to model complex systems: providing evidence of their applicability to controlled systems is a significant step forward, broadening the application field of combinatorial methods and disclosing opportunities for hybrid off-line/on-line optimization.

An efficient routing protocol for internet of medical things focusing hot spot node problem

The healthcare budget is increasing day-by-day as the population of the world increases. The same is the case regarding the workload of health care workers, that is, doctors and other practitioners. Under such a scenario where workload and cost are increasing drastically, there is a dire need of integrating recent technological enhancements with the said domain. Since the last decade, a lot of work is in the process considering the said integration bringing revolutionary changes. For remote monitoring, existing systems use different types of Internet of things devices that measure different health parameters. One of the major problems in such a system is to find an optimum routing approach that can resolve energy and thermal issues that are taking the limelight in the research arena. In this article, a dynamic routing technique is proposed which is keen to connect multiple in vivo/ex vivo Internet of things devices and a sink (focusing thermal and energy problem) and then forwarding data from sink to remote location for monitoring. Performance parameters are kept energy efficiency and thermal awareness and analytical results show that the proposed protocol supersedes existing approaches in said metrics.

THERMAL AWARE LINK ENERGY EFFICIENT SCHEME FOR BODY AREA NETWORKS

A daunting task in Wireless Body Area Networks (WBANs) is still to develop Effective routing techniques. Small-sized nodes are installed on or within the human body to monitor human health conditions which then deliver the data to servers for analysis. During sensing and data transfer, biomedical sensors work continuously and the temperature of the nodes may rise beyond the threshold limit. This temperature rise may damage the human body tissues as well as the routing mechanism in terms of path losses. To keep the temperature at its normal working value, a priority-based selection of routes is required to prevent data loss during transmission. This will ensure safe and accurate data delivery at the destination. A protocol called “Thermal Aware Link Energy Efficient Scheme for WBANs” (TALEEBA) for workers is proposed to monitor the health of workers in factories. One of the four sinks will collect the data of the nearest worker in the field. As the body temperature of any worker is detected to rise, an alarm will be generated and the supervisor of the workplace will ask the worker to be replaced by some other worker. The same mechanism will continue till the task ends. Our proposed TALEEBA (Thermal Aware Link Energy Efficient Scheme for WBANs) scheme is aligned with current LAEEBA and THE-FAME WBAN schemes. In simulations, we analyze our protocol in terms of stability period, network lifetime, residual energy, a packet sent, packet dropped, and throughput. Hence, the results show stability and network life 50%, a packet sent 20% and throughput 23% are improved in comparison with LABEEA and THE-FAME protocols.

High Throughput and Thermal Aware Routing Protocol (HTTRP) for Wireless Body Area Networks

High Throughput and Thermal Aware Routing Protocol (HTTRP) for Wireless Body Area Networks

On Fundamental Principles for Thermal-Aware Design on Periodic Real-Time Multi-Core Systems

With the exponential rise of the transistor count in one chip, the thermal problem has become a pressing issue in computing system design. While there have been extensive methods and techniques published for design optimization with thermal awareness, there is a need for more rigorous and formal thermal analysis in designing real-time systems and applications that demand a strong exception guarantee. In this article, we analytically prove a series of fundamental properties and principles concerning the RC thermal model, peak temperature identification, and peak temperature reduction for periodic real-time systems, which are general enough to be applied on 2D and 3D multi-core platforms. These findings enhance the worst-case temperature predictability in runtime scenarios, as well as help to develop more effective thermal management policy, which is key to thermal-constrained periodic real-time system design.

An Improved Matrix Generation Framework for Thermal Aware Placement in VLSI

Since hotspots and temperature gradients are reliability and performance-critical issues in processors, thermal awareness finds a vital place in the processor design cycle. Incorporating thermal awareness at the level of physical design, this work proposes a new, fast, and efficient thermal aware placement algorithm called the Thermal Aware Matrix Placement Optimizer (TAMPO) for gate arrays. The algorithm TAMPO is composed of the following components: an improved heat diffusion aware cell arrangement technique called the Initial Matrix Generator (IMaGe), a unique stochastic thermal model based on a thermally improved interpretation of the well known Matrix Synthesis Problem (MSP) and a Simulated Annealing (SA) engine for finding the global optimum solution. TAMPO targets to reduce the peak temperature while maintaining improved values of temperature gradients and the standard deviation in cell temperature with respect to the average chip temperature. This work also presents a methodology, the Co-optimized TAMPO, which extends the concept of TAMPO to simultaneously optimize the thermal attributes and the wirelength of a chip. The proposed algorithms realize a placement in matrix arrangement and upon experimentation on the ISCAS89 benchmark circuits encouraging results have been obtained.

Thermal Aware Task Scheduling for Enhanced Cyber-Physical Systems Sustainability

Cyber-Physical Systems (CPS) are increasingly used in a variety of transportation, healthcare, electricity grid, and other applications. Thermal stress is often a major concern for processors embedded in such systems. High operating temperatures can dramatically shorten processor life. This, in turn, can require provisioning of significant amounts of additional computational hardware to withstand more frequent failures, with obvious implications for sustainability. This paper describes a novel approach to reduce thermally-induced damage in CPS processors by targeting Dynamic Voltage and Frequency Scaling (DVFS) to high-activity task phases. That is, by preferentially slowing down high-activity task phases, significant additional savings in energy and thermal stress can be attained for a given amount of computational slowdown; this approach is shown to be superior to conventional methods that use DVFS without regard to activity levels. Also, our proposed task reassignment across cores is driven by estimates of current core reliability, which is superior to the usual approach of simply using either current temperature or temperature history. Our approach leads to a significant reliability improvement (around 20 percent) over baseline DVFS techniques.

STREAM: Stress and Thermal Aware Reliability Management for 3-D ICs

Accurate and fast reliability management is important for 3-D integrated circuits (3-D ICs) because of the severe on-chip thermal and reliability problems. However, due to the lack of stress information and difficulties in implementing management method for reliability, existing full-chip reliability management methods suffer from low management accuracy and high system performance degradation. In this paper, we propose a new stress and thermal aware reliability management method for 3-D ICs called STREAM. Unlike traditional methods which do not perform explicit stress analysis due to the large computing cost, STREAM employs an artificial neural network-based stress model to estimate stress accurately at runtime. In order to further improve the reliability management accuracy and improve the system performance, a lifetime estimator with lifetime banking technology and a specially designed lifetime model predictive control are integrated into the reliability management framework. Our numerical results show that STREAM performs the stress and thermal aware full-chip reliability management with both high accuracy and speed. It is able to boost the performance of 3-D ICs and outperforms the state-of-the-art 3-D IC reliability management method.

A Survey on Thermal Aware Routing Protocols in Wireless Body Area Network

A Survey on Thermal Aware Routing Protocols in Wireless Body Area Network

Dynamic Task Mapping and Scheduling with Temperature-Awareness on Network-on-Chip based Multicore Systems

Abstract Technology scaling has enabled integration of a large number of intellectual property cores, memory units and processing elements on a single chip. This has led to the evolution of multicore systems. In recent years, these systems have gained popularity as execution platforms for real-time applications. However, due to the close spacing of the cores and uneven heat dissipation among the processors of the multicore platform, thermal hotspots are formed. Ensuring thermal safety in real-time systems is a challenging task, especially when various applications are executed with different timing constraints. In real-time dynamic systems, applications unknown at design-time, are submitted by users at runtime. In order to perform on-the-fly task assignment to the processors of the multicore platform, different dynamic resource allocation approaches are reported in the literature. However, most of the dynamic task allocation approaches lack in thermal awareness and do not consider thermal safety of the chip. Also, most of the existing thermal-aware task mapping methods rely on offline thermal analysis which is either unavailable in dynamic scenarios or is computationally too expensive for providing end-to-end thermal-aware solution at runtime. In this work, we propose an improved algorithm for dynamic thermal-aware task mapping and scheduling using a non-profiling based strategy for NoC based multi-core systems. By using a combination of threshold based thermal management scheme and dynamic task (re)allocation, the proposed thermal-aware approach helps to maintain the thermal safety of the chip. Simulation results demonstrate that the proposed algorithm achieves 25.3% and 45.6% reduction in chip peak temperature compared to recent dynamic task allocation approaches ATM [1] and DEAMS [2] respectively. Also, the proposed thermal-aware strategy satisfies the task deadline while reducing the average packet delay of the allocated applications. When compared to other thermal-aware methods such as ATM [1] and TAPP [3], the proposed strategy improves the deadline satisfaction of the tasks by 37.8% and 67.3% respectively.