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:20220715T000515Z
LOCATION:Level 2 Lobby
DTSTART;TZID=America/Los_Angeles:20220712T180000
DTEND;TZID=America/Los_Angeles:20220712T190000
UID:dac_DAC 2022_sess264_WIP138@linklings.com
SUMMARY:You Already Have It: A Generator-Free Low-Precision DNN Training F
 ramework using Stochastic Rounding
DESCRIPTION:Networking Reception, Work-in-Progress Poster\n\nYou Already H
 ave It: A Generator-Free Low-Precision DNN Training Framework using Stocha
 stic Rounding\n\nYuan, Chang, Jin, Lu, Li...\n\nTo generate random numbers
  for use in the stochastic rounding technique, a large number of random nu
 mber generators are used in conventional designs.\nTherefore, extra hardwa
 re costs are introduced inevitably.\nIn this paper, we innovatively propos
 e to employ the stochastic property of the neural network training process
  and directly extract random numbers from neural networks in a self-suffic
 ient manner. \nWe explore the quality of the obtained random numbers from 
 different sources and methods, and propose a generator-free framework for 
 low-precision neural network training.\nWe validate the effectiveness of o
 ur framework on a variety of deep learning tasks.
END:VEVENT
END:VCALENDAR