2nd ROAD4NN Workshop: Research Open Automatic Design for Neural Networks
Co-located with the 58th Design Automation Conference (DAC 2021), Dec 5th, 2021
In the past decade, machine learning, especially neural network based deep learning, has achieved amazing success. Various types of neural networks, such as CNNs, RNNs, LSTMs, BERT, GNNs, SNNs, and the recent vision transformers, have been deployed for various industrial applications like image classification, speech recognition, natural language understanding, autonomous driving, and automated control. On one hand, there is a very fast algorithm evolvement of neural network models. Almost every week there is a new model from a major academic and/or industry institute. On the other hand, all major industry giants have been developing and/or deploying specialized hardware platforms to accelerate the performance and improve the energy-efficiency of neural networks across the cloud and edge devices. This includes Nvidia GPUs, Google TPUs, ARM and Qualcomm mobile CPUs and GPUs, programmable DSPs and NPUs, Intel Nervana/Habana/Loihi ASICs, Xilinx and Intel FPGAs, Microsoft Brainwave, Amazon Inferentia, to name just a few. However, there is a significant gap between the fast algorithm evolvement and staggering hardware development, hence calling for broader participation in software-hardware co-design from both academia and industry.
In this workshop, we focus on the open research of automated design for neural networks, a holistic open source approach to general-purpose computer systems broadly inspired by neural networks. More specifically, we discuss full stack open source infrastructure support to develop and deploy novel neural networks, including novel algorithms and applications, hardware architectures and emerging digital/analog devices, as well as programming, compiler, system, and tool support. We plan to bring together academic and industry experts to share their experience, discuss challenges they face as well as potential focus areas for the community.
We are soliciting work-in-progress papers from the community. Workshop topics include, but are not limited to:
· New algorithm advancement of neural networks
· Bio-plausible neural network models
· Neural network model compression, quantization, and network architecture search
· Application of neural networks into new areas
· Hardware acceleration and architecture for neural networks
· New analog and mixed-signal circuits and architecture for neural networks
· Abstraction to bridge the algorithm and hardware gap for neural networks
· Compilation and design automation support to map neural networks to hardware platforms
· System support to deploy neural networks in cloud and edge devices
· Benchmarks for various neural network models and hardware accelerators
· Other research infrastructures that enable the above studies
Submission guidelines:
Interested authors are encouraged to submit their work-in-progress papers (up to four pages) through EasyChair (link: https://easychair.org/conferences/?conf=road4nn2021). Authors are encouraged to submit preliminary work for new projects and early results. All papers will be reviewed with a double-blind review process. Manuscripts should not exceed 4 single-spaced, double-column pages using 10-point size font on 8.5×11 inch pages (IEEE conference style), including references, figures, and tables. All papers must be submitted electronically in PDF format. Accepted papers will be invited to give a 25 mins talk with 5 min Q/A for each talk; there will be no proceedings, so that authors can still submit their papers to other conferences/journals. The deadline for submission is Aug 8, 11:59 PM (Pacific Time), 2021.
Student presentation award:
We will select the three Best Student Presentations from the workshop and send a mysterious gift to each student awardee.
Important dates:
· Paper submission: Aug 8, 2021, 11:59 PM (Pacific Time)
· Author notification: Aug 31, 2021 (Pacific Time)
· Workshop date: Dec 5, 2021 (Pacific Time)
Organizers:
· Zhenman Fang, Simon Fraser University, Canada (zhenman@sfu.ca)
· Yanzhi Wang, Northeastern University, US (yanz.wang@northeastern.edu)
· Zhe Chen, UCLA, US (zhechen@ucla.edu)
Past ROAD4NN workshop:
We have organized the first ROAD4NN workshop co-located with the Design Automation Conference 2020 (DAC 2020) on Jul 19, 2020. It is a huge success with more than 500 registrations from more than 200 academic institutions and industrial companies. For the first workshop, it is invitation only, and we have invited a number of renowned researchers to give talks, who come from academia and industry and some of them have successful experience in co-founding machine learning startup companies. For example, we have invited Prof. Deming Chen from UIUC to give a keynote. Other invited researchers include Xilinx Fellow Ashish Sirasao, Microsoft Brainwave project researcher Fanny Nina Paravecino (female), Prof. Song Han from MIT, Prof. Zhiru Zhang from Cornell, Prof. Zhangyang (Atlas) Wang from TAMU, Prof. Yingyan Lin from Rice (female), Prof. Peng Li from UCSB, to name just a few. The full schedule (with talk slides) can be found on our ROAD4NN 2020 workshop website: https://sites.google.com/view/road4nn/schedule.