A length adaptive algorithm-hardware co-design of transformer on FPGA through sparse attention and dynamic pipelining

Hongwu Peng, Shaoyi Huang, Shiyang Chen, Bingbing Li, Tong Geng, Ang Li, Weiwen Jiang, Wujie Wen, Jinbo Bi, Hang Liu, Caiwen Ding. A length adaptive algorithm-hardware co-design of transformer on FPGA through sparse attention and dynamic pipelining. In Rob Oshana, editor, DAC '22: 59th ACM/IEEE Design Automation Conference, San Francisco, California, USA, July 10 - 14, 2022. pages 1135-1140, ACM, 2022. [doi]

@inproceedings{PengHCLGLJWBLD22,
  title = {A length adaptive algorithm-hardware co-design of transformer on FPGA through sparse attention and dynamic pipelining},
  author = {Hongwu Peng and Shaoyi Huang and Shiyang Chen and Bingbing Li and Tong Geng and Ang Li and Weiwen Jiang and Wujie Wen and Jinbo Bi and Hang Liu and Caiwen Ding},
  year = {2022},
  doi = {10.1145/3489517.3530585},
  url = {https://doi.org/10.1145/3489517.3530585},
  researchr = {https://researchr.org/publication/PengHCLGLJWBLD22},
  cites = {0},
  citedby = {0},
  pages = {1135-1140},
  booktitle = {DAC '22: 59th ACM/IEEE Design Automation Conference, San Francisco, California, USA, July 10 - 14, 2022},
  editor = {Rob Oshana},
  publisher = {ACM},
  isbn = {978-1-4503-9142-9},
}