Castelló, A.; SERGIO BARRACHINA; Dolz Zaragozá, MF.; Enrique S. Quintana-Ortí; San Juan-Sebastian, P.; Tomás Domínguez, AE. (2022). High performance and energy efficient inference for deep learning on multicore ARM processors using general optimization techniques and BLIS. Journal of Systems Architecture. 125:1-9. https://doi.org/10.1016/j.sysarc.2022.102459
Por favor, use este identificador para citar o enlazar este ítem: http://hdl.handle.net/10251/197569
Title:
|
High performance and energy efficient inference for deep learning on multicore ARM processors using general optimization techniques and BLIS
|
Author:
|
Castelló, Adrián
SERGIO BARRACHINA
DOLZ ZARAGOZÁ, MANUEL FRANCISCO
Enrique S. Quintana-Ortí
San Juan-Sebastian, Pablo
Tomás Domínguez, Andrés Enrique
|
UPV Unit:
|
Universitat Politècnica de València. Departamento de Informática de Sistemas y Computadores - Departament d'Informàtica de Sistemes i Computadors
Universitat Politècnica de València. Escola Tècnica Superior d'Enginyeria Informàtica
|
Issued date:
|
|
Abstract:
|
[EN] We evolve PyDTNN, a framework for distributed parallel training of Deep Neural Networks (DNNs), into an efficient inference tool for convolutional neural networks. Our optimization process on multicore ARM processors ...[+]
[EN] We evolve PyDTNN, a framework for distributed parallel training of Deep Neural Networks (DNNs), into an efficient inference tool for convolutional neural networks. Our optimization process on multicore ARM processors involves several high-level transformations of the original framework, such as the development and integration of Cython routines to exploit thread-level parallelism; the design and development of micro-kernels for the matrix multiplication, vectorized with ARM's NEON intrinsics, that can accommodate layer fusion; and the appropriate selection of several cache configuration parameters tailored to the memory hierarchy of the target ARM processors.Our experiments evaluate both inference throughput (measured in processed images/s) and inference latency (i.e., time-to-response) as well as energy consumption per image when varying the level of thread parallelism and the processor power modes. The experiments with the new inference engine are reported for the ResNet50 v1.5 model on the ImageNet dataset from the MLPerf suite using the ARM v8.2 cores in the NVIDIA Jetson AGX Xavier board. These results show superior performance compared with the well-spread TFLite from Google and slightly inferior results when compared with ArmNN, the native library from ARM for DNN inference.
[-]
|
Subjects:
|
Convolutional neural network
,
Inference
,
Multicore low-power processors
|
Copyrigths:
|
Reconocimiento - No comercial - Sin obra derivada (by-nc-nd)
|
Source:
|
Journal of Systems Architecture. (issn:
1383-7621
)
|
DOI:
|
10.1016/j.sysarc.2022.102459
|
Publisher:
|
Elsevier
|
Publisher version:
|
https://doi.org/10.1016/j.sysarc.2022.102459
|
Project ID:
|
info:eu-repo/grantAgreement/AEI/Plan Estatal de Investigación Científica y Técnica y de Innovación 2013-2016/TIN2017-82972-R/ES/TECNICAS ALGORITMICAS PARA COMPUTACION DE ALTO RENDIMIENTO CONSCIENTE DEL CONSUMO ENERGETICO Y RESISTENTE A ERRORES/
info:eu-repo/grantAgreement/GVA//PROMETEO%2F2019%2F109//COMUNICACION Y COMPUTACION INTELIGENTES Y SOCIALES/
info:eu-repo/grantAgreement/GVA//CDEIGENT%2F2018%2F014//Plan GenT/
|
Thanks:
|
This research was partially sponsored by projects TIN2017-82972-R of Ministerio de Ciencia, Innovacion y Universidades, Spain and Prometeo/2019/109 of the Generalitat Valenciana, Spain. Adrian Castello was supported by the ...[+]
This research was partially sponsored by projects TIN2017-82972-R of Ministerio de Ciencia, Innovacion y Universidades, Spain and Prometeo/2019/109 of the Generalitat Valenciana, Spain. Adrian Castello was supported by the Juan de la Cierva-Formacion project FJC2019-039222-I of the Ministerio de Ciencia, Innovacion y Universidades, Spain. Manuel F. Dolz was also supported by the Plan GenT project CDEIGENT/2018/014 of the Generalitat Valenciana, Spain.
[-]
|
Type:
|
Artículo
|