FTRANS: Energy-efficient acceleration of transformers using FPGA

Bingbing Li, Santosh Pandey, Haowen Fang, Yanjun Lyv, Ji Li, Jieyang Chen, Mimi Xie, Lipeng Wan, Hang Liu, Caiwen Ding

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

78 Scopus citations

Abstract

In natural language processing (NLP), the "Transformer"architecture was proposed as the first transduction model replying entirely on self-Attention mechanisms without using sequence-Aligned recurrent neural networks (RNNs) or convolution, and it achieved significant improvements for sequence to sequence tasks. The introduced intensive computation and storage of these pre-Trained language representations has impeded their popularity into computation and memory constrained devices. The field-programmable gate array (FPGA) is widely used to accelerate deep learning algorithms for its high parallelism and low latency. However, the trained models are still too large to accommodate to an FPGA fabric. In this paper, we propose an efficient acceleration framework, Ftrans, for transformer-based large scale language representations. Our framework includes enhanced block-circulant matrix (BCM)-based weight representation to enable model compression on large-scale language representations at the algorithm level with few accuracy degradation, and an acceleration design at the architecture level. Experimental results show that our proposed framework significantly reduce the model size of NLP models by up to 16 times. Our FPGA design achieves 27.07× and 81 × improvement in performance and energy efficiency compared to CPU, and up to 8.80× improvement in energy efficiency compared to GPU.

Original languageEnglish
Title of host publicationProceedings of the ACM/IEEE International Symposium on Low Power Electronics and Design, ISLPED 2020
PublisherAssociation for Computing Machinery
ISBN (Electronic)9781450370530
DOIs
StatePublished - Aug 10 2020
Event2020 ACM/IEEE International Symposium on Low Power Electronics and Design, ISLPED 2020 - Virtual, Online, United States
Duration: Aug 10 2020Aug 12 2020

Publication series

NameACM International Conference Proceeding Series

Conference

Conference2020 ACM/IEEE International Symposium on Low Power Electronics and Design, ISLPED 2020
Country/TerritoryUnited States
CityVirtual, Online
Period08/10/2008/12/20

Fingerprint

Dive into the research topics of 'FTRANS: Energy-efficient acceleration of transformers using FPGA'. Together they form a unique fingerprint.

Cite this