SyNERGY: An energy measurement and prediction framework for Convolutional Neural Networks on Jetson TX1Citation formats

Standard

SyNERGY: An energy measurement and prediction framework for Convolutional Neural Networks on Jetson TX1. / Rodrigues, Crefeda; Riley, Graham; Luján, Mikel.

PDPTA'18 - The 24th International Conference on Parallel and Distributed Processing Techniques and Applications . 2018.

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Harvard

Rodrigues, C, Riley, G & Luján, M 2018, SyNERGY: An energy measurement and prediction framework for Convolutional Neural Networks on Jetson TX1. in PDPTA'18 - The 24th International Conference on Parallel and Distributed Processing Techniques and Applications . <https://csce.ucmss.com/cr/books/2018/ConferenceReport?ConferenceKey=PDP>

APA

Rodrigues, C., Riley, G., & Luján, M. (Accepted/In press). SyNERGY: An energy measurement and prediction framework for Convolutional Neural Networks on Jetson TX1. In PDPTA'18 - The 24th International Conference on Parallel and Distributed Processing Techniques and Applications https://csce.ucmss.com/cr/books/2018/ConferenceReport?ConferenceKey=PDP

Vancouver

Rodrigues C, Riley G, Luján M. SyNERGY: An energy measurement and prediction framework for Convolutional Neural Networks on Jetson TX1. In PDPTA'18 - The 24th International Conference on Parallel and Distributed Processing Techniques and Applications . 2018

Author

Rodrigues, Crefeda ; Riley, Graham ; Luján, Mikel. / SyNERGY: An energy measurement and prediction framework for Convolutional Neural Networks on Jetson TX1. PDPTA'18 - The 24th International Conference on Parallel and Distributed Processing Techniques and Applications . 2018.

Bibtex

@inproceedings{1d3ec2f85cd044dfa404087a5249e56d,
title = "SyNERGY: An energy measurement and prediction framework for Convolutional Neural Networks on Jetson TX1",
abstract = "There is a huge demand for on-device execution of deep learning algorithms on mobile and embedded platforms. These devices present constraints on the application due to limited hardware resources and power. However, current evaluation studies in existing deep learning frameworks (for example, Caffe, Tensorflow, Torch and others) are limited to performance measurements of these applications on high-end CPUs and GPUs. In this work, we propose{"}SyNERGY{"} a fine-grained energy measurement (that is, at specific layers) and prediction framework for deep neural networks on embedded platforms. We integrate ARM{\textquoteright}s Streamline Performance Analyser with standard deep learningframeworks such as Caffe and CuDNNv5 to quantify the energy-use of deep convolutional neural networks on the Nvidia Jetson Tegra X1. Our measurement framework provides an accurate breakdown of actual energy consumptionand performance across all layers in the neural network while our prediction framework models the energy-use in terms of target-specific performance counters such as SIMD and bus accesses and application specific parameters such as Multiply and Accumulate (MAC) counts. Our experimental results using 9 representative Deep Convolutional Neural Network shows that a multi-variable linear regression model based on hardware performance counters alone achieves an average prediction test error of 8.0 5.96% compared to actual energy measurements. Surprisingly, we find that it is possible to refine the model to predict the number of SIMD instructions and main memory accesses solely fromthe application{\textquoteright}s Multiply-Accumulate (MAC) counts with an average prediction test error of 0.81 0.77% and 17.09 13% respectively. This alleviates the need for actual measurements giving a final average prediction test error of 7.0 6.0% using solely the application{\textquoteright}s MAC counts as input.",
keywords = "Energy Measurement, Energy prediction, Convolutional Neural Networks, Mobile computing",
author = "Crefeda Rodrigues and Graham Riley and Mikel Luj{\'a}n",
year = "2018",
month = jun,
day = "24",
language = "English",
isbn = "1-60132-487-1",
booktitle = "PDPTA'18 - The 24th International Conference on Parallel and Distributed Processing Techniques and Applications",

}

RIS

TY - GEN

T1 - SyNERGY: An energy measurement and prediction framework for Convolutional Neural Networks on Jetson TX1

AU - Rodrigues, Crefeda

AU - Riley, Graham

AU - Luján, Mikel

PY - 2018/6/24

Y1 - 2018/6/24

N2 - There is a huge demand for on-device execution of deep learning algorithms on mobile and embedded platforms. These devices present constraints on the application due to limited hardware resources and power. However, current evaluation studies in existing deep learning frameworks (for example, Caffe, Tensorflow, Torch and others) are limited to performance measurements of these applications on high-end CPUs and GPUs. In this work, we propose"SyNERGY" a fine-grained energy measurement (that is, at specific layers) and prediction framework for deep neural networks on embedded platforms. We integrate ARM’s Streamline Performance Analyser with standard deep learningframeworks such as Caffe and CuDNNv5 to quantify the energy-use of deep convolutional neural networks on the Nvidia Jetson Tegra X1. Our measurement framework provides an accurate breakdown of actual energy consumptionand performance across all layers in the neural network while our prediction framework models the energy-use in terms of target-specific performance counters such as SIMD and bus accesses and application specific parameters such as Multiply and Accumulate (MAC) counts. Our experimental results using 9 representative Deep Convolutional Neural Network shows that a multi-variable linear regression model based on hardware performance counters alone achieves an average prediction test error of 8.0 5.96% compared to actual energy measurements. Surprisingly, we find that it is possible to refine the model to predict the number of SIMD instructions and main memory accesses solely fromthe application’s Multiply-Accumulate (MAC) counts with an average prediction test error of 0.81 0.77% and 17.09 13% respectively. This alleviates the need for actual measurements giving a final average prediction test error of 7.0 6.0% using solely the application’s MAC counts as input.

AB - There is a huge demand for on-device execution of deep learning algorithms on mobile and embedded platforms. These devices present constraints on the application due to limited hardware resources and power. However, current evaluation studies in existing deep learning frameworks (for example, Caffe, Tensorflow, Torch and others) are limited to performance measurements of these applications on high-end CPUs and GPUs. In this work, we propose"SyNERGY" a fine-grained energy measurement (that is, at specific layers) and prediction framework for deep neural networks on embedded platforms. We integrate ARM’s Streamline Performance Analyser with standard deep learningframeworks such as Caffe and CuDNNv5 to quantify the energy-use of deep convolutional neural networks on the Nvidia Jetson Tegra X1. Our measurement framework provides an accurate breakdown of actual energy consumptionand performance across all layers in the neural network while our prediction framework models the energy-use in terms of target-specific performance counters such as SIMD and bus accesses and application specific parameters such as Multiply and Accumulate (MAC) counts. Our experimental results using 9 representative Deep Convolutional Neural Network shows that a multi-variable linear regression model based on hardware performance counters alone achieves an average prediction test error of 8.0 5.96% compared to actual energy measurements. Surprisingly, we find that it is possible to refine the model to predict the number of SIMD instructions and main memory accesses solely fromthe application’s Multiply-Accumulate (MAC) counts with an average prediction test error of 0.81 0.77% and 17.09 13% respectively. This alleviates the need for actual measurements giving a final average prediction test error of 7.0 6.0% using solely the application’s MAC counts as input.

KW - Energy Measurement

KW - Energy prediction

KW - Convolutional Neural Networks

KW - Mobile computing

M3 - Conference contribution

SN - 1-60132-487-1

BT - PDPTA'18 - The 24th International Conference on Parallel and Distributed Processing Techniques and Applications

ER -