← All publications
DATE, 2026LLM / VLA

Boosting the Performance of Tree-Based Speculative Decoding of LLMs on FPGAs

Tielong Liu, Gang Li, Zitao Mo, Zeyu Zhu, Minnan Pei and Jian Cheng

Overview

FPGA acceleration of tree-based speculative decoding for large language models.

面向 FPGA 的树状推测解码性能优化与大语言模型推理加速。

  • Speculative decoding
  • FPGA
  • LLM inference

DATE, 2026. Public paper and code links will be added when available.

Research by Minnan Pei (裴旻楠) and collaborators · Email