Evaluating the impact of job scheduling and power management on processor lifetime for chip multiprocessors

Coskun, Ayse K.; Strong, Richard; Tullsen, Dean M.; Rosing, Tajana

doi:10.1145/1555349.1555369

Cited by 89 publications

(64 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use several most common temperature induced intrinsic hard failure mechanisms, including electromigration, time dependent dielectric breakdown, thermal cycling and disk failure [48], [18], [45], to analyze the reliability degradation caused by a thermal attack. We use λ to represent the failure rate of each failure mechanism.…”

Section: E Damage Analysismentioning

confidence: 99%

See 1 more Smart Citation

Reduced Cooling Redundancy: A New Security Vulnerability in a Hot Data Center

Gao

Wang

et al. 2018

Proceedings 2018 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

Abstract-Data centers have been growing rapidly in recent years to meet the surging demand of cloud services. However, the expanding scale and powerful servers generate a great amount of heat, resulting in significant cooling costs. A trend in modern data centers is to raise the temperature and maintain all servers in a relatively hot environment. While this can save on cooling costs given benign workloads running in servers, the hot environment increases the risk of cooling failure. In this paper, we unveil a new vulnerability of existing data centers with aggressive cooling energy saving policies. Such a vulnerability might be exploited to launch thermal attacks that could severely worsen the thermal conditions in a data center. Specifically, we conduct thermal measurements and uncover effective thermal attack vectors at the server, rack, and data center levels. We also present damage assessments of thermal attacks. Our results demonstrate that thermal attacks can (1) largely increase the temperature of victim servers degrading their performance and reliability, (2) negatively impact on thermal conditions of neighboring servers causing local hotspots, (3) raise the cooling cost, and (4) even lead to cooling failures. Finally, we propose effective defenses to prevent thermal attacks from becoming a serious security threat to data centers.

show abstract

Section: E Damage Analysismentioning

confidence: 99%

“…They reported that temperature is strongly correlated with disk failures. Coskun et al [18] studied the reliability of different job scheduling and power management methods. They proposed a fine grained technique to simulate the thermal behaviors and input the thermal trace to the reliability model.…”

Section: Related Workmentioning

confidence: 99%

Reduced Cooling Redundancy: A New Security Vulnerability in a Hot Data Center

Gao

Wang

et al. 2018

Proceedings 2018 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

show abstract

“…There are a few recent approaches aimed at resource management for either power consumption and/or lifetime optimization in symmetric [1,5,11,4,14,6] or asymmetric [10,21,18] architectures. The authors in [4] propose an off-line technique to improve the lifetime reliability of MPSoCs, while [11] describes a combination of design-time and runtime techniques to optimize it.…”

Section: Introduction and Related Workmentioning

confidence: 99%

“…It is commonly agreed upon that online adaptation is essential when dealing with aging, temperature, or energy related optimization, due to the lack of significant information that could drive a design-time solution space exploration. Thus, when moving to on-line optimization, [1] is one of the first approaches considering both lifetime reliability and energy consumption in MPSoCs; however, the two dimensions are optimized separately. Energy and reliability optimization is considered in [5] as well: the proposed hybrid approach does not consider computational energy optimization and DVFS-enabled architectures.…”

Section: Introduction and Related Workmentioning

confidence: 99%

Combined on-line lifetime-energy optimization for asymmetric multicores

Bolchini

Carminati

Mitra

et al. 2016

2016 IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems (DFT)

View full text Add to dashboard Cite

Abstract-In this paper we present an architectural and online resource management solution to optimize lifetime reliability of asymmetric multicores while minimizing the system energy consumption, targeting both single nodes (multicores) as well as multiple ones (cluster of multicores). The solution exploits the different characteristics of the computing resources to achieve the desired performance while optimizing the lifetime/energy tradeoff. The experimental results show that a combined optimization of energy and lifetime allows for achieving an extended lifetime (similar to the one pursued by lifetime-only optimization solutions) with a marginal energy consumption detriment (less than 2%) with respect to energy-aware but aging-unaware systems.

show abstract

“…However, the DVFS technique reduces the temperature by sacrificing the computing system performance. Based on the DVFS technique, a lot of researches have been published [31,58,60]. They either tried to analyze the trade off between the performance and power and thermal dissipation, or focused on optimizing the thermal issue by reducing processor peak temperature.…”

Section: Dynamic Voltage and Frequency Scalingmentioning

confidence: 99%

Practical Dynamic Thermal Management on Intel Desktop Computer

Quan

Liu

View full text Add to dashboard Cite

Professor Gang Quan, Major ProfessorFueled by increasing human appetite for high computing performance, semiconductor technology has now marched into the deep sub-micron era. As transistor size keeps shrinking, more and more transistors are integrated into a single chip. This has increased tremendously the power consumption and heat generation of IC chips. The rapidly growing heat dissipation greatly increases the packaging/cooling costs, and adversely affects the performance and reliability of a computing system. In addition, it also reduces the processor's life span and may even crash the entire computing system. Therefore, dynamic thermal management (DTM) is becoming a critical problem in modern computer system design.Extensive theoretical research has been conducted to study the DTM problem.However, most of them are based on theoretically idealized assumptions or simplified models. While these models and assumptions help to greatly simplify a complex problem and make it theoretically manageable, practical computer systems and applications must deal with many practical factors and details beyond these models or assumptions.The goal of our research was to develop a test platform that can be used to validate theoretical results on DTM under well-controlled conditions, to identify the limitations of existing theoretical results, and also to develop new and practical DTM techniques. This dissertation details the background and our research efforts in this v endeavor. Specifically, in our research, we first developed a customized test platform based on an Intel desktop. We then tested a number of related theoretical works and examined their limitations under the practical hardware environment. With these limitations in mind, we developed a new reactive thermal management algorithm for single-core computing systems to optimize the throughput under a peak temperature constraint. We further extended our research to a multicore platform and developed an effective proactive DTM technique for throughput maximization on multicore processor based on task migration and dynamic voltage frequency scaling technique. The significance of our research lies in the fact that our research complements the current extensive theoretical research in dealing with increasingly critical thermal problems and enabling the continuous evolution of high performance computing systems.

show abstract

Evaluating the impact of job scheduling and power management on processor lifetime for chip multiprocessors

Cited by 89 publications

References 29 publications

Reduced Cooling Redundancy: A New Security Vulnerability in a Hot Data Center

Reduced Cooling Redundancy: A New Security Vulnerability in a Hot Data Center

Combined on-line lifetime-energy optimization for asymmetric multicores

Practical Dynamic Thermal Management on Intel Desktop Computer

Contact Info

Product

Resources

About