Fine-Grain Power Breakdown of Modern Out-of-Order Cores and Its Implications on Skylake-Based Systems

Haj-Yihia, Jawad; Yasin, Ahmad; Asher, Yosi Ben; Mendelson, Avi

doi:10.1145/3018112

Cited by 33 publications

(21 citation statements)

References 18 publications

(30 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Li et al [33] present Multicore Power Area and Timing simulator (McPAT) to estimate the power consumption for various components in a multiprocessor, which includes shared caches, integrated memory controllers, in-order and out-of-order processor cores and networks-on-chip. Haj-Yihia et al [34] proposed a linear power predictive model for Intel Skylake based CPUs based on selected PMCs that are highly positively correlated with power consumption. Mair et al [35] presented Manila which is a power model based on PMC space generated as densely populated points gathered via a large number of synthetic applications.…”

Section: Software Based Energy Predictive Modelsmentioning

confidence: 99%

See 1 more Smart Citation

A Comparative Study of Methods for Measurement of Energy of Computing

et al. 2019

View full text Add to dashboard Cite

Energy of computing is a serious environmental concern and mitigating it is an important technological challenge. Accurate measurement of energy consumption during an application execution is key to application-level energy minimization techniques. There are three popular approaches to providing it: (a) System-level physical measurements using external power meters; (b) Measurements using on-chip power sensors and (c) Energy predictive models. In this work, we present a comprehensive study comparing the accuracy of state-of-the-art on-chip power sensors and energy predictive models against system-level physical measurements using external power meters, which we consider to be the ground truth. We show that the average error of the dynamic energy profiles obtained using on-chip power sensors can be as high as 73% and the maximum reaches 300% for two scientific applications, matrix-matrix multiplication and 2D fast Fourier transform for a wide range of problem sizes. The applications are executed on three modern Intel multicore CPUs, two Nvidia GPUs and an Intel Xeon Phi accelerator. The average error of the energy predictive models employing performance monitoring counters (PMCs) as predictor variables can be as high as 32% and the maximum reaches 100% for a diverse set of seventeen benchmarks executed on two Intel multicore CPUs (one Haswell and the other Skylake). We also demonstrate that using inaccurate energy measurements provided by on-chip sensors for dynamic energy optimization can result in significant energy losses up to 84%. We show that, owing to the nature of the deviations of the energy measurements provided by on-chip sensors from the ground truth, calibration can not improve the accuracy of the on-chip sensors to an extent that can allow them to be used in optimization of applications for dynamic energy. Finally, we present the lessons learned, our recommendations for the use of on-chip sensors and energy predictive models and future directions.

show abstract

Section: Software Based Energy Predictive Modelsmentioning

confidence: 99%

“…Techniques that use expert advice or intuition to pick a subset of PMCs and that, in experts' opinion, are dominant contributors to energy consumption [34].…”

mentioning

confidence: 99%

A Comparative Study of Methods for Measurement of Energy of Computing

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Haj-Yihia et al [10] start with a set of 23 PMCs (offered by Likwid) based on expert knowledge of the Intel architecture. Then they perform linear regression iteratively where they drop PMCs (one by one) that do not impact the average prediction error of their model.…”

Section: Techniques For Selection Of Pmcs For Energy Predictive Modelingmentioning

confidence: 99%

“…It is noteworthy that some non-additive PMCs are used as predictor variables in many energy predictive models [5,6,10,20,23]. These are ICache events, L2 Transactions, and L2 Requests.…”

Section: Additivity Of Likwid Pmcsmentioning

confidence: 99%

“…In the case study, they use 35 carefully selected PMCs (out of a total of 390 available in the platform) in their linear regression model for predicting dynamic energy consumption. [1,9,10] select PMCs manually, based on in-depth study of architecture and empirical analysis. [8,15,18,21,22,27,30] select PMCs that are highly correlated with energy consumption using Spearman's rank correlation coefficient (or Pearson's correlation coefficient) and principal component analysis (PCA).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Additivity: A Selection Criterion for Performance Events for Reliable Energy Predictive Modeling

Shahid

Fahad

Reddy

et al. 2017

JSFI

View full text Add to dashboard Cite

Performance events or performance monitoring counters (PMCs) are now the dominant predictor variables for modeling energy consumption. Modern hardware processors provide a large set of PMCs. Determination of the best subset of PMCs for energy predictive modeling is a nontrivial task given the fact that all the PMCs can not be determined using a single application run. Several techniques have been devised to address this challenge. While some techniques are based on a statistical methodology, some use expert advice to pick a subset (that may not necessarily be obtained in one application run) that, in experts' opinion, are significant contributors to energy consumption. However, the existing techniques have not considered a fundamental property of predictor variables that should have been applied in the first place to remove PMCs unfit for modeling energy. We address this oversight in this paper.We propose a novel selection criterion for PMCs called additivity, which can be used to determine the subset of PMCs that can potentially be used for reliable energy predictive modeling. It is based on the experimental observation that the energy consumption of a serial execution of two applications is the sum of energy consumptions observed for the individual execution of each application. A linear predictive energy model is consistent if and only if its predictor variables are additive in the sense that the vector of predictor variables for a serial execution of two applications is the sum of vectors for the individual execution of each application. The criterion, therefore, is based on a simple and intuitive rule that the value of a PMC for a serial execution of two applications is equal to the sum of its values obtained for the individual execution of each application. The PMC is branded as non-additive on a platform if there exists an application for which the calculated value differs significantly from the value observed for the application execution on the platform. The use of non-additive PMCs in a model renders it inconsistent.We study the additivity of PMCs offered by the popular state-of-the-art tools, Likwid and PAPI, by employing a detailed experimental methodology on a modern Intel Haswell multicore server CPU. We show that many PMCs in Likwid and PAPI that are widely used in models as key predictor variables are non-additive. This brings into question the reliability and the reported prediction accuracy of these models.

show abstract