SpecMamba: Accelerating Mamba Inference on FPGA with Speculative Decoding

出版物
In International Conference on Computer-Aided Design (ICCAD)
Linfeng Zhong
PhD Student (co-advised w/ Prof. Yuan Wang, 2026)
Songqiang Xu
PhD, Tsinghua University

2+ first-author papers

Huifeng Wen
Master Student (2024)
Tong Xie
PhD Student (co-advised w/ Prof. Runsheng Wang, 2024)
Qingyu Guo
Senior Software Engineer@ByteDance

1st author papers x 4, 1st Place IEEE AICAS Contest

李萌
李萌
助理教授

我通过算法—硬件协同设计,研究高效且保护隐私的人工智能系统,并将其应用于大语言模型、具身智能和多模态智能。