A Probabilistic Graphical Model-based Approach for Minimizing Energy Under Performance Constraints

Mishra, Nikita; Zhang, Huazhe; Lafferty, John; Hoffmann, Henry

doi:10.1145/2694344.2694373

Cited by 46 publications

(11 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A hybrid combination of offline and online techniques has recently been proposed to minimize the energy consumption under a performance constraint [25]. This technique employs probabilistic graphical models to estimate the power and performance for unknown applications at runtime based on previously stored offline application data.…”

Section: Related Researchmentioning

confidence: 99%

An Online Learning Methodology for Performance Modeling of Graphics Processors

Gupta

Babu²,

Ayoub³

et al. 2018

IEEE Trans. Comput.

View full text Add to dashboard Cite

Approximately 18% of the 3.2 million smartphone applications rely on integrated graphics processing units (GPUs) to achieve competitive performance. Graphics performance, typically measured in frames per second, is a strong function of the GPU frequency, which in turn has a significant impact on mobile processor power consumption. Consequently, dynamic power management algorithms have to assess the performance sensitivity to the frequency accurately to choose the operating frequency of the GPU effectively. Since the impact of GPU frequency on performance varies rapidly over time, there is a need for online performance models that can adapt to varying workloads. This paper presents a light-weight adaptive runtime performance model that predicts the frame processing time of graphics workloads at runtime without apriori characterization. We employ this model to estimate the frame time sensitivity to the GPU frequency, i.e., the partial derivative of the frame time with respect to the GPU frequency. The proposed model does not rely on any parameter learned offline. Our experiments on the Intel Minnowboard MAX platform running common GPU benchmarks show that the mean absolute percentage error in frame time and frame time sensitivity prediction are 4.2% and 6.7%, respectively.

show abstract

Section: Related Researchmentioning

confidence: 99%

An Online Learning Methodology for Performance Modeling of Graphics Processors

Gupta

Babu²,

Ayoub³

et al. 2018

IEEE Trans. Comput.

View full text Add to dashboard Cite

show abstract

“…Imes et al [16] used controller theory and linear programming to find energy-optimized configurations for an application with soft real-time constraints at runtime. Mishra et al [24] used hierarchical Bayesian model in machine learning to find energy-optimized configurations. Snowdon et al [26] developed a power management framework called Koala which models the energy consumption of the platform and monitors an application' energy behavior.…”

Section: Related Work -Overview Of Energy Modelsmentioning

confidence: 99%

“…Significant efforts have been devoted to developing power and energy models in literature [2,8,7,17,18,16,24,26]. However, there are no analytic models for multithreaded algorithms that are both applicable to a wide range of algorithms and comprehensively validated yet (cf.…”

Section: Introductionmentioning

confidence: 99%

“…• Devising a new general energy model ICE for analyzing the energy com- [24] No General Yes POET [16] No General Yes Koala [26] No General Yes Roofline [8,7] No General Yes Energy scalability [17,18] Yes General No Sequential energy complexity [25] No General Yes Alonso et al [2] Yes Algorithm-specific Yes Malossi et al [23] Yes Algorithm-specific Yes ICE model (this study) Yes General Yes…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

ICE: A General and Validated Energy Complexity Model for Multithreaded Algorithms

Tran

2016

2016 IEEE 22nd International Conference on Parallel and Distributed Systems (ICPADS)

View full text Add to dashboard Cite

Like time complexity models that have significantly contributed to the analysis and development of fast algorithms, energy complexity models for parallel algorithms are desired as crucial means to develop energy efficient algorithms for ubiquitous multicore platforms. Ideal energy complexity models should be validated on real multicore platforms and applicable to a wide range of parallel algorithms. However, existing energy complexity models for parallel algorithms are either theoretical without model validation or algorithm-specific without ability to analyze energy complexity for a wide-range of parallel algorithms. This paper presents a new general validated energy complexity model for parallel (multithreaded) algorithms. The new model abstracts away possible multicore platforms by their static and dynamic energy of computational operations and data access, and derives the energy complexity of a given algorithm from its work, span and I/O complexity. The new model is validated by different sparse matrix vector multiplication (SpMV) algorithms and dense matrix multiplication (matmul) algorithms running on high performance computing (HPC) platforms (e.g., Intel Xeon and Xeon Phi). The new energy complexity model is able to characterize and compare the energy consumption of SpMV and matmul kernels according to three aspects: different algorithms, different input matrix types and different platforms. The prediction of the new model regarding which algorithm consumes more energy with different inputs on different platforms, is confirmed by the experimental results. In order to improve the usability and accuracy of the new model for a wide range of platforms, arXiv:1605.08222v2 [cs.DC] 4 Oct 2016 the platform parameters of ICE model are provided for eleven platforms including HPC, accelerator and embedded platforms.

show abstract

“…The FLNM is designed based on the dependence among the high-level features [32], noisy labels and true labels [25] and can effectively capture the noise distribution. A probabilistic graphical model (PGM) [33] is used to describe the dependence structure of the noise estimating model. The four nodes of the underlying graph represent the input image, high-level features, true labels and noisy labels.…”

Section: Introductionmentioning

confidence: 99%

Object Extraction From Very High-Resolution Images Using a Convolutional Neural Network Based on a Noisy Large-Scale Dataset

Cheng

et al. 2019

IEEE Access

View full text Add to dashboard Cite

In recent years, convolutional neural networks (CNNs) have made great achievements in object extraction from very high-resolution (VHR) images. However, most existing approaches require large quantities of clean and accurate training data to achieve impressive classification results. The presence of inaccurate labels in training datasets is known to deteriorate the performance of CNNs. In this paper, we introduce a novel efficient method for improving the robustness when training CNN on the dataset with relatively noisy labels. First, we propose a feature and label noise model (FLNM) to model the noisy label distribution in the training dataset. Then, we use a multitask deep learning framework (MDLF) to integrate the FLNM into the training process of CNN. Finally, a novel loss function concerning the high-level features is introduced to efficiently train the MDLF. We evaluate our method on datasets from Massachusetts and compare this method with other state-of-the-art methods. The experimental results demonstrate the effectiveness of the proposed method in improving the classification performance of CNNs trained with noisy training dataset.

show abstract

A Probabilistic Graphical Model-based Approach for Minimizing Energy Under Performance Constraints

Cited by 46 publications

References 70 publications

An Online Learning Methodology for Performance Modeling of Graphics Processors

An Online Learning Methodology for Performance Modeling of Graphics Processors

ICE: A General and Validated Energy Complexity Model for Multithreaded Algorithms

Object Extraction From Very High-Resolution Images Using a Convolutional Neural Network Based on a Noisy Large-Scale Dataset

Contact Info

Product

Resources

About