TY - GEN
T1 - A 47.4µJ/epoch trainable deep convolutional neural network accelerator for in-situ personalization on smart devices
AU - Choi, Seungkyu
AU - Sim, Jaehyeong
AU - Kang, Myeonggu
AU - Choi, Yeongjae
AU - Kim, Hyeonuk
AU - Kim, Lee Sup
N1 - Publisher Copyright:
© 2019 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.
PY - 2019/11
Y1 - 2019/11
N2 - A scalable deep learning accelerator supporting both inference and training is implemented for device personalization of deep convolutional neural networks. It consists of three processor cores operating with distinct energy-efficient dataflow for different types of computation in CNN training. Two cores conduct forward and backward propagation in convolutional layers and utilize a masking scheme to reduce 88.3% of intermediate data to store for training. The third core executes weight update process in convolutional layers and inner product computation in fully connected layers with a novel large window dataflow. The system enables 8-bit fixed point datapath with lossless training and consumes 47.4µJ/epoch for a customized deep CNN model.
AB - A scalable deep learning accelerator supporting both inference and training is implemented for device personalization of deep convolutional neural networks. It consists of three processor cores operating with distinct energy-efficient dataflow for different types of computation in CNN training. Two cores conduct forward and backward propagation in convolutional layers and utilize a masking scheme to reduce 88.3% of intermediate data to store for training. The third core executes weight update process in convolutional layers and inner product computation in fully connected layers with a novel large window dataflow. The system enables 8-bit fixed point datapath with lossless training and consumes 47.4µJ/epoch for a customized deep CNN model.
UR - http://www.scopus.com/inward/record.url?scp=85108985295&partnerID=8YFLogxK
U2 - 10.1109/A-SSCC47793.2019.9056972
DO - 10.1109/A-SSCC47793.2019.9056972
M3 - Conference contribution
AN - SCOPUS:85108985295
T3 - Proceedings - 2019 IEEE Asian Solid-State Circuits Conference, A-SSCC 2019
SP - 57
EP - 60
BT - Proceedings - 2019 IEEE Asian Solid-State Circuits Conference, A-SSCC 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 15th IEEE Asian Solid-State Circuits Conference, A-SSCC 2019
Y2 - 4 November 2019 through 6 November 2019
ER -