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:20211208T164511Z
LOCATION:
DTSTART;TZID=America/Los_Angeles:20211205T080000
DTEND;TZID=America/Los_Angeles:20211205T170000
UID:dac_DAC 2021_sess190_WKSHP103@linklings.com
SUMMARY:MOVED TO VIRTUAL: 2nd ROAD4NN Workshop: Research Open Automatic De
 sign for Neural Networks
DESCRIPTION:Workshop\n\nMOVED TO VIRTUAL: 2nd ROAD4NN Workshop: Research O
 pen Automatic Design for Neural Networks\n\nFang, Wang, Chen\n\nIn the pas
 t decade, machine learning, especially neural network based deep learning,
  has achieved an amazing success. Various neural networks, such as CNNs, R
 NNs, LSTMs, BERT, GNNs, and SNNs, have been deployed for various industria
 l applications like image classification, speech recognition, and automate
 d control. On one hand, there is a very fast algorithm evolvement of neura
 l network models, almost every week there is a new model from a major acad
 emic and/or industry institute. On the other hand, all major industry gian
 ts have been developing and/or deploying specialized hardware platforms to
  accelerate the performance and energy-efficiency of neural networks acros
 s the cloud and edge devices. This include Nvidia GPU, Intel Nervana/Haban
 a/Loihi ASICs, Xilinx FPGA, Google TPU, Microsoft Brainwave, Amazon Infere
 ntia, to name just a few. However, there is a significant gap between the 
 fast algorithm evolvement and staggering hardware development, hence calli
 ng for broader participation in software-hardware co-design from both acad
 emia and industry.\n\nIn this workshop, we focus on the research open auto
 matic design for neural networks, a holistic open source approach to gener
 al-purpose computer systems broadly inspired by neural networks. More spec
 ifically, we discuss full stack open source infrastructure support to deve
 lop and deploy novel neural networks, including novel algorithms and appli
 cations, hardware architectures and emerging devices, as well as programmi
 ng, system, and tool support. We plan to bring together academic and indus
 try experts to share their experience, discuss challenges they face as wel
 l as potential focus areas for the community. Below is the planed workshop
  content.\n\nFirst, we will solicit work-in-progress papers (four pages) f
 rom the community and accepted papers will be published in proceedings (op
 tional) and be invited to give a 25 mins talk with 5 min Q/A for each talk
 . Workshop topics include, but are not limited to:\n•	New algorithm 
 advancement of neural networks\n•	Bio-plausible neural network model
 s\n•	Neural network model compression and quantization\n•	Appl
 ication of neural networks into new areas\n•	Hardware acceleration a
 nd architecture for neural networks\n•	New circuits and devices for 
 neural networks\n•	Abstraction to bridge the algorithm and hardware 
 gap for neural networks\n•	Compilation and design automation support
  to map neural networks to hardware platforms\n•	System support to d
 eploy neural networks in cloud and edge devices\n•	Benchmarks for va
 rious neural network models and hardware accelerators\n•	Other resea
 rch infrastructures that enable the above studies\n\nSecond, we plan to in
 vite a few established researchers both from academic (e.g., MIT, UCLA, Co
 rnell, Stanford, Duke, Tsinghua University, Peking University) and/or indu
 stry (e.g., Intel, Xilinx, Nvidia, Microsoft, Google) to give a keynote an
 d invited talks.\n\nThird, we plan to organize a panel session to give the
  audience more Q/A time. We plan to invite some established researchers fr
 om the aforementioned institutes for the panel.\n\nVirtual Program: Presen
 ted In-Person
END:VEVENT
END:VCALENDAR
