Synthesis of workload monitors for on-line stress prediction

Baranowski, Rafał; Cook, Alejandro; Imhof, Michael E.; Liu, Chang; Wunderlich, Hans-Joachim

doi:10.1109/dft.2013.6653596

Cited by 11 publications

(4 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The temperature monitor can help to detect temperature increases due to developing faults in the core, which is usually the case. This could be combined later with a processor workload monitor [8]. The delay monitors, in combination of voltage monitors, can show the system (frequency) operating degradation that can be caused by aging behaviour.…”

Section: The Health-monitoring Infrastructurementioning

confidence: 99%

Design of an Embedded Health Monitoring Infrastructure for Accessing Multi-processor SoC Degradation

Zhao

Kerkhoff

2014

2014 17th Euromicro Conference on Digital System Design

View full text Add to dashboard Cite

An embedded health-monitoring infrastructure for a highly reliable MP-SoC for data-streaming systems is presented. Different from the traditional approach of a dependable design, our infrastructure is based on life-time prognostics from healthmonitoring sensors that are embedded near the target processor. This enables the preventive repair by spare parts or priority ranking of tasks among processors. Focus of this paper is mainly the health-monitoring scheme and the sensor structure with simulation results.

show abstract

Section: The Health-monitoring Infrastructurementioning

confidence: 99%

Design of an Embedded Health Monitoring Infrastructure for Accessing Multi-processor SoC Degradation

Zhao

Kerkhoff

2014

2014 17th Euromicro Conference on Digital System Design

View full text Add to dashboard Cite

show abstract

“…The VT‐sensitive RO can also support a direct NBTI monitor. Although the accurate NBTI degradation depends on the switching behaviour of each target device [17], holistic degradation of chip performance caused by NBTI would correlate with the VT count. The authors in [10] discuss that coarse‐grained monitors can be more practical than fine‐grained monitors in some applications like balancing workload among identical multi‐cores.…”

Section: Stress Monitor In Comparison With Prior Workmentioning

confidence: 99%

“…The downside is the loss of target generality. The stress monitor is the indirect type, which monitors the environmental conditions such as T or V indirectly [17]. Since T and sometimes V in addition are common causes of various wear‐out, the indirect type can cope with various wear‐out targets, including EM, or stress‐migration (SM).…”

Section: Stress Monitor In Comparison With Prior Workmentioning

confidence: 99%

Wear‐out stress monitor utilising temperature and voltage sensitive ring oscillators

Kan

Shimada

Okagaki

et al. 2018

IET Circuits, Devices & Systems

View full text Add to dashboard Cite

The authors propose an on-chip wear-out monitoring technique, which is based on monitoring the environmental conditions experienced by a digital circuit. The frequency of the T-sensitive ring oscillator (RO) emulates the wear-out stress strength caused by the temperature conditions based on the model of exponential dependence of the stress on the inverse of temperatures. The frequency of the VT-sensitive RO emulates the stress due to time-dependent dielectric breakdown, which is stressed by voltages as well as temperatures. Thus, the accumulated counts driven by the ROs directly indicate the total wearout stress that the product has experienced so far. The measured results of a test chip fabricated by 28 nm High-k Metal Gate process confirm the expected dependence of T-/VT-sensitive RO frequencies on temperatures and voltages, enabling the emulation of wear-out. The methodology is presented to estimate the stress amount of various wear-out factors having different thermal activation energies. The proposed wear-out stress monitor would make automotive microcontrollers more reliable when they operate at boosted voltages and elevated temperatures to meet performance requirements of cutting-edge applications such as advanced driver assistance systems.

show abstract

“…In principle, workload monitors can be synthesized using the technique presented in [9]. Our experimental results show, however, that this approach results in prohibitive area overhead that often exceeds 100% for the usual number of RCGs required for accurate aging rate prediction.…”

Section: Workload Monitoringmentioning

confidence: 99%

On-Line Prediction of NBTI-induced Aging Rates

Baranowski¹,

Firouzi²,

Kiamehr³

et al. 2015

Design, Automation &Amp; Test in Europe Conference &Amp; Exhibition (DATE), 2015

View full text Add to dashboard Cite

Nanoscale technologies are increasingly susceptible to aging processes such as Negative-Bias Temperature Instability (NBTI) which undermine the reliability of VLSI systems. Existing monitoring techniques can detect the violation of safety margins and hence make the prediction of an imminent failure possible. However, since such techniques can only detect measurable degradation effects which appear after a relatively long period of system operation, they are not well suited to early aging prediction and proactive aging alleviation. This work presents a novel method for the monitoring of NBTI-induced degradation rate in digital circuits. It enables the timely adoption of proper mitigation techniques that reduce the impact of aging. The developed method employs machine learning techniques to find a small set of so called Representative Critical Gates (RCG), the workload of which is correlated with the degradation of the entire circuit. The workload of RCGs is observed in hardware using so called workload monitors. The output of the workload monitors is evaluated on-line to predict system degradation experienced within a configurable (short) period of time, e.g. a fraction of a second. Experimental results show that the developed monitors predict the degradation rate with an average error of only 1.6% at 4.2% area overhead.

show abstract

Synthesis of workload monitors for on-line stress prediction

Cited by 11 publications

References 22 publications

Design of an Embedded Health Monitoring Infrastructure for Accessing Multi-processor SoC Degradation

Design of an Embedded Health Monitoring Infrastructure for Accessing Multi-processor SoC Degradation

Wear‐out stress monitor utilising temperature and voltage sensitive ring oscillators

On-Line Prediction of NBTI-induced Aging Rates

Contact Info

Product

Resources

About