Full-chip runtime error-tolerant thermal estimation and prediction for practical thermal management

Wang, Hong; Tan,; Liao,; Quintanilla,; Gupta, Rajiv

doi:10.1109/iccad.2011.6105408

Cited by 16 publications

(7 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared to the previous solution, they use the power dissipated by each core improving the estimation during the transient, and moreover it is capable to filter out the noise that is usually present in the thermal sensors readings. Similarly Wang et al [32] starting from a detailed thermal model of the device assume that cores temperature does not change between two consecutive time intervals and use this assumption to improve robustness of the thermal model to errors in the power input and thermal model parameters. It does not consider noise in the thermal sensors readings (i.e.…”

Section: Related Workmentioning

confidence: 98%

An Effective Gray-Box Identification Procedure for Multicore Thermal Modeling

Beneventi

Bartolini

Tilli

et al. 2014

IEEE Trans. Comput.

View full text Add to dashboard Cite

Aggressive thermal management is a critical feature for high-end computing platforms, as worst-case thermal budgeting is becoming unaffordable. Reactive thermal management, which sets temperature thresholds to trigger thermal capping actions, is too "nearsighted," and it may lead to severe performance degradation and thermal overshoots. More aggressive proactive thermal managements minimize performance penalty with smooth optimal control. These techniques require knowledge of thermal models, which have to be accurate and simple to make the controls effective, while keeping their complexity limited. In practice, these models are not provided by manufacturers, and in most cases, they strongly depend on the deployment environment. Hence, procedures to automatically derive thermal models in the field are needed. In this paper, we propose a gray-box procedure to learn a compact and physically consistent model for multicore chips. We leverage the physical consistency of the proposed model to tame the model complexity and to face large quantization noise in measurements. We exploit Output Error structures along with Levenberg-Marquardt and Least Squares optimization algorithms. We tackle the problem in a real-life contest: we developed a complete infrastructure for model building and thermal data collection in the Linux environment, and we tested it on an Intel Nehalem-based server CPU.

show abstract

Section: Related Workmentioning

confidence: 98%

An Effective Gray-Box Identification Procedure for Multicore Thermal Modeling

Beneventi

Bartolini

Tilli

et al. 2014

IEEE Trans. Comput.

View full text Add to dashboard Cite

show abstract

“…Firstly, the 5 thermal error sequences in Figure 7 are converted into 3 series in Figures 8(a)-10(a) based on (27)- (28). The spindle axial elongation time series in thedirection was directly from the measured data, and the radial thermal yaw and pitch angle series , , , were obtained by applying (27) and (28), respectively. Then the Augmented Dickey-Fuller (ADF) Test Algorithm was applied to identify the stationarity of the thermal error sequences , , , , , and the calculation showed that thermal elongation and angles were nonstationary series.…”

Section: Thermal Errors Prediction and Compensationmentioning

confidence: 99%

“…So we can forecast its trends and make necessary control on it. Wang et al applied time series analysis method to establish a spindle thermal error Mathematical Problems in Engineering model and compensated errors [25][26][27], and they acquired better results. This paper focuses on the spindle system of a box-type precision CNC coordinate boring machine.…”

Section: Introductionmentioning

confidence: 99%

Thermal-Induced Errors Prediction and Compensation for a Coordinate Boring Machine Based on Time Series Analysis

Yang¹,

Zhang

Bao³

et al. 2014

Mathematical Problems in Engineering

View full text Add to dashboard Cite

To improve the CNC machine tools precision, a thermal error modeling for the motorized spindle was proposed based on time series analysis, considering the length of cutting tools and thermal declined angles, and the real-time error compensation was implemented. A five-point method was applied to measure radial thermal declinations and axial expansion of the spindle with eddy current sensors, solving the problem that the three-point measurement cannot obtain the radial thermal angle errors. Then the stationarity of the thermal error sequences was determined by the Augmented Dickey-Fuller Test Algorithm, and the autocorrelation/partial autocorrelation function was applied to identify the model pattern. By combining both Yule-Walker equations and information criteria, the order and parameters of the models were solved effectively, which improved the prediction accuracy and generalization ability. The results indicated that the prediction accuracy of the time series model could reach up to 90%. In addition, the axial maximum error decreased from 39.6 μm to 7 μm after error compensation, and the machining accuracy was improved by 89.7%. Moreover, theX/Y-direction accuracy can reach up to 77.4% and 86%, respectively, which demonstrated that the proposed methods of measurement, modeling, and compensation were effective.

show abstract

“…But thermal effects and the resulting reliability concerns, such as NBTI effect, are influenced by the placement of CPU cores and shared caches, loading of programs, and cooling solutions at the package level. So it is vital to accurately estimate the temperature during the floorplanning, architecture design [Huang et al 2004;Skadron et al 2003;Huang et al 2006;Li et al 2009Li et al , 2008Yang et al 2007] and the runtime [Wang et al 2011] of the multicore microprocessors.…”

Section: Introductionmentioning

confidence: 99%

Composable thermal modeling and simulation for architecture-level thermal designs of multicore microprocessors

Wang

Tan

et al. 2013

ACM Trans. Des. Autom. Electron. Syst.

Self Cite

View full text Add to dashboard Cite

Efficient temperature estimation is vital for designing thermally efficient, lower power and robust integrated circuits in nanometer regime. Thermal simulation based on the detailed thermal structures no longer meets the demanding tasks for efficient design space exploration. The compact and composable model-based simulation provides a viable solution to this difficult problem. However, building such thermal models from detailed thermal structures was not well addressed in the past. In this article, we propose a new compact thermal modeling technique, called ThermComp, standing for thermal modeling with composable modules. ThermComp can be used for fast thermal design space exploration for multicore microprocessors. The new approach builds the composable model from detailed structures for each basic module using the finite difference method and reduces the model complexity by the sampling-based model order reduction technique. These composable models are then used to assemble different multicore architecture thermal models and realized into SPICE-like netlists. The resulting thermal models can be simulated by the general circuit simulator SPICE. ThermComp tries to preserve the accuracy of fine-grained models with the speed of coarse-grained models. Experimental results on a number of multicore microprocessor architectures show the new approach can easily build accurate thermal systems from compact composable models for fast architecture thermal analysis and optimization and is much faster than the existing HotSpot method with similar accuracy.

show abstract

Full-chip runtime error-tolerant thermal estimation and prediction for practical thermal management

Cited by 16 publications

References 19 publications

An Effective Gray-Box Identification Procedure for Multicore Thermal Modeling

An Effective Gray-Box Identification Procedure for Multicore Thermal Modeling

Thermal-Induced Errors Prediction and Compensation for a Coordinate Boring Machine Based on Time Series Analysis

Composable thermal modeling and simulation for architecture-level thermal designs of multicore microprocessors

Contact Info

Product

Resources

About