FPGA-based accelerator for losslessly quantized convolutional neural networks

Mankit Sit, Ryosuke Kazami, Hideharu Amano

研究成果: Conference contribution

9 被引用数 (Scopus)

抄録

Convolutional Neural Networks (CNN) have been widely used for various computer vision tasks. While GPUs are the most common platform for CNN implementation, FPGAs are promising alternatives to provide better energy efficiency. Recent work demonstrates the potential of network quantization to reduce the model size and enhance computation efficiency while maintaining comparable accuracy to the full precision counterparts. Quantized CNN is especially suitable for FPGA implementation due to the presence of values with non-trivial bitwidth. In this paper, we present the design of an FPGA-based accelerator for losslessly quantized CNNs using High Level Synthesis tool. The experiment result shows that our design achieves 12.9 GOPS/Watt for quantized Alexnet on Imagnet Dataset.

本文言語English
ホスト出版物のタイトル2017 International Conference on Field-Programmable Technology, ICFPT 2017
出版社Institute of Electrical and Electronics Engineers Inc.
ページ295-298
ページ数4
ISBN(電子版)9781538626559
DOI
出版ステータスPublished - 2018 2月 2
イベント16th IEEE International Conference on Field-Programmable Technology, ICFPT 2017 - Melbourne, Australia
継続期間: 2017 12月 112017 12月 13

出版物シリーズ

名前2017 International Conference on Field-Programmable Technology, ICFPT 2017
2018-January

Other

Other16th IEEE International Conference on Field-Programmable Technology, ICFPT 2017
国/地域Australia
CityMelbourne
Period17/12/1117/12/13

ASJC Scopus subject areas

  • ハードウェアとアーキテクチャ
  • ソフトウェア
  • 器械工学

フィンガープリント

「FPGA-based accelerator for losslessly quantized convolutional neural networks」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル