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:20211208T164407Z
LOCATION:3018
DTSTART;TZID=America/Los_Angeles:20211207T141500
DTEND;TZID=America/Los_Angeles:20211207T143800
UID:dac_DAC 2021_sess116_RESEARCH583@linklings.com
SUMMARY:FIXAR: A Fixed-Point Deep Reinforcement Learning Platform with Qua
 ntization-Aware Training and Adaptive Parallelism
DESCRIPTION:Research Manuscript\n\nFIXAR: A Fixed-Point Deep Reinforcement
  Learning Platform with Quantization-Aware Training and Adaptive Paralleli
 sm\n\nYang\n\nIn this paper, we present a deep reinforcement learning plat
 form named FIXAR which employs fixed-point data types and arithmetic units
  for the first time using a SW/HW codesign approach. Starting from 32-bit 
 fixed-point data, Quantization-Aware Training (QAT) reduces its data preci
 sion based on the range of activations and performs retraining to minimize
  the reward degradation. FIXAR proposes the adaptive array engine composed
  of 16x16 configurable processing elements to support both intra-layer par
 allelism and intra-batch parallelism for high-throughput inference and tra
 ining. Finally, FIXAR was implemented on Xilinx U50 and achieves 2638.0 IP
 S/W, which is 15.4 times more energy efficient than GPU.\n\nTopic: Design\
 n\nKeyword: AI/ML System Design\n\nVirtual Program: Presented In-Person
END:VEVENT
END:VCALENDAR