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:20240626T180033Z
LOCATION:Level 2 Lobby
DTSTART;TZID=America/Los_Angeles:20240625T180000
DTEND;TZID=America/Los_Angeles:20240625T190000
UID:dac_DAC 2024_sess236_RESEARCH020@linklings.com
SUMMARY:Athena: Add More Intelligence to RMT-based Network Data Plane with
  Low-bit Quantization
DESCRIPTION:Work-in-Progress Poster\n\nYunkun Liao, Hanyue Lin, Jingya Wu,
  Wenyan Lu, Xiaowei Li, and Guihai Yan (State Key Lab of Processors, Insti
 tute of Computing Technology, Chinese Academy of Sciences)\n\nPerforming p
 er-packet Neural Network (NN) inference on the network data plane is requi
 red for high-quality and fast decision-making in computer networking. Howe
 ver, data plane architecture like the Reconfigurable Match Tables (RMT) pi
 peline has limited support for NN. Previous efforts have utilized Binary N
 euron Networks (BNNs) as a compromise, but the accuracy loss of BNN is hig
 h. Inspired by the accuracy gain of the two-bit model. this paper proposes
  Athena. Athena can deploy the sparse low-bit quantization (two-bit and fo
 ur-bit) model on RMT. Compared with the BNN-based state-of-the-art, Athena
  is cost-effective regarding accuracy loss reduction, inference latency, a
 nd chip area overhead.\n\nTopic: AI, Autonomous Systems, Cloud, Design, ED
 A, Embedded Systems, IP, Security
END:VEVENT
END:VCALENDAR