BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260730T152637Z
LOCATION:
DTSTART;TZID=America/Los_Angeles:20260728T123700
DTEND;TZID=America/Los_Angeles:20260728T123700
UID:dac_DAC 2026_sess315_RESEARCH2753@linklings.com
SUMMARY:An Energy-Efficient Dataflow Architecture for Efficient MoE Model 
 Inference
DESCRIPTION:Kexin Li, Qinggang Wang, Chuanhui Qi, Shaoxian Xu, and Wenkan 
 Huang (Huazhong University of Science and Technology); Xiangzheng Yang (Zh
 engzhou University); ruoshi li (Huazhong University of Science and Technol
 ogy); Bo Liu (Zhengzhou University); and Long Zheng, Xiaofei Liao, and Hai
  Jin (Huazhong University of Science and Technology)\n\nThe Mixture-of-Exp
 erts (MoE) model sparsely activates a subset of experts for each token, ca
 using each expert to process a varying number of tokens during inference. 
 However, existing GPU-based MoE inference frameworks adopt a fixed tensor 
 size for the inputs to each expert, requiring padding when an expert recei
 ves fewer tokens, which leads to substantial redundant computation.\nIn re
 sponse, we propose a token-stationary dataflow that uniformly abstracts th
 e multiplication between an expert's parameters and input tensors with var
 ying token counts. Based on this dataflow, we design a reconfigurable syst
 olic array that eliminates padding-incurred redundant computation. Evaluat
 ion demonstrates that our design outperforms the state-of-the-art GPU-base
 d MoE inference frameworks significantly.\n\nTopics: Quantum\n\nTrack: DES
 6. Quantum Computing, Systems\n\nSession Chairs: Hokeun Kim (Arizona State
  University); BaekGyu Kim (Daegu Gyeongbuk Institute of Science and Techno
 logy); Xiang Chen (Peking University); Sangyeob Kim (Yonsei University); R
 obert M. Radway (University of Pennsylvania); Prashanth Mohan (University 
 of Pittsburgh); Xiaoxuan Yang (University of Virginia, Stanford University
 ); Ramtin Zand (University of South Carolina); Hadi Kamali (University of 
 Central Florida); Jiafeng Xie (Villanova University); Prabuddha Chakrabort
 y (University of Maine); Junichiro Kadomoto (University of Tokyo); and Yos
 uke Ueno (RIKEN)\n\n
END:VEVENT
END:VCALENDAR