The Workshop Schedule
November 10-11, 2020
Day 1: Keynotes and Panel Sessions
Opening and introduction: 10: 00am - 10: 30am EST
Workshop opening:
Joydip Kundu, Acting Deputy Assistant Director of the CISE Directorate at National Science Foundation
Keynote talk 1: 10: 30am - 11: 15am EST
Talk title: Machine learning hardware from a DARPA / MTO perspective
Break: 10mins
Panel 1: 11: 25am - 12:40pm EST
Scalable/Efficient hardware architectures
Moderator: Jan M. Rabaey (UC Berkeley)
Panelists: Kailash Gopalakrishnan (IBM), Dmitri Vainbrand (Nvidia), Joseph Hassoun (Samsung Semiconductor), R. Iris Bahar (Brown University), and Christos Kozyrakis (Stanford University)
Scriber: Yakun Sophia Shao (UC Berkeley)
Keynote talk 2: 12: 40pm - 1: 25pm EST
Talk title: Domain-specific hardware accelerators for deep learning
Break: 30mins
Keynote talk 3: 1: 55pm - 2: 40pm EST
Talk title: Technology integration enables energy-efficient on-device intelligence
Panel 2: 2: 40pm - 3: 55pm EST
Supporting algorithms, tools, and benchmarks
Moderator: Hadi Esmaeilzadeh (UC San Diego)
Panelists: Vijay Janapa Reddi (Harvard University), Sachin Sapatnekar (University of Minnesota), Luis Ceze (University of Washington), and Zhiru Zhang (Cornell University)
Scriber: Tushar Krishna (Georgia Tech)
Break: 10mins
Panel 3: 4: 05pm - 5: 20pm EST
Enabling technologies
Moderator: Hai "Helen" Li (Duke University)
Panelists: Suman Datta (University of Notre Dame), Shimeng Yu (Georgia Tech), Takashi Ando (IBM), and Sayeef Salahuddin (UC Berkeley)
Scriber: Priyadarshini Panda (Yale University)
Day 2: Keynotes and Breakout Sessions
Keynote talk 4: 10: 05am - 10: 50am EST
Talk title: Integrated photonics for high performance computing
Break: 10mins
Breakout sessions: 11: 00am - 12: 30pm EST
Breakout 1: Algorithm-hardware co-design
Moderator: Bryan Jacobs (DARPA)
Panelists: Song Han (MIT), Deming Chen (Univ. of Illinois at Urbana-Champaign), Farinaz Koushanfar (UC San Diego), Mike Davies (Intel), and Wei Wen (Facebook)
Scriber: Bo Yuan (Rutgers University)
Breakout 2: The role of neuromorphic/non-digital computing
Moderator: Shantanu Chakrabartty (Washington Univ. in St. Louis) and Yiran Chen (Duke University)
Panelists: Gert Cauwenberghs (UC San Diego), Rahul Sarpeshkar (Dartmouth College), Tobi Delbruck (ETH Zurich), Qinru Qiu (Syracuse University), Keshab Parhi (University of Minnesota), and Wei Lu (University of Michigan)
Scriber: Fan Chen (Indiana University Bloomington)
Break: 30mins
Keynote talk 5: 1: 00pm - 1: 45pm EST
Talk title: How to evaluate efficient deep neural network approaches
Breakout sessions: 1: 45pm - 3: 15pm EST
Breakout 3: Driving needs from the end-users/applications
Moderator: Keith Rebello (DARPA)
Panelists: Aadeel Akhtar (PSYONIC), Edwin Park (Qualcomm), Chiao Liu (Facebook), Chris Dick (Nvidia), and Peter Kairouz (Google)
Scriber: Yuhao Zhu (University of Rochester)
Breakout 4: Opportunities for Photonics in ML Systems
Moderator: Kelvin Wagner (Univ. of Colorado Boulder) and Volker Sorger (George Washington University)
Panelists: Liang Feng (University of Pennsylvania), Stefan Preble (Rochester Institute of Technology), Manya Ghobadi (MIT), and Mo Li (University of Washington)
Scriber: Nathan Youngblood (University of Pittsburgh)
Break: 10mins
Keynote talk 6: 3: 25pm - 4: 10pm EST
Talk title: Scientific questions to fuel neural nets and TPU design
