Local Activity Principle : The Elan Vital of Complexity and Life

Speaker Biosketch

Leon O. Chua (Life Fellow, IEEE) is a Professor Emeritus and Research Professor of UC Berkeley. He is widely known for his invention of the Memristor. His research has been recognized through 17 honorary doctorates from major universities in Europe and Japan, and holds seven U.S. patents. He was conferred numerous prestigious awards, including the first IEEE Gustav Robert Kirchhoff Award, the Guggenheim Fellow, the 2019 EDS Celebrated Member Prize—The highest recognition of the IEEE Electron Devices Society, and the 2020 Julius Springer Prize in applied physics. He was inducted to be a Foreign Member of the European Academy of Sciences (Academia Europaea) in 1997 and the Hungarian Academy of Sciences in 2007. He was elected as Confrerie des Chevaliers du Tastevin in 2000. When not immersed in Science, he relaxes by searching for Wagner’s leitmotifs, musing over Kandinsky’s chaos, and contemplating Wittgenstein’s inner thoughts. He obtained his B.S.E.E. degree from the Mapua Institute of Technology, Philippines, M.S.E.E. degree from MIT, and Ph.D. degree from the University of Illinois at Urbana-Champaign.

Abstract

      In this talk I will review several unsolved deep scientific problems that many luminaries, including Galvani, Boltzmann, Schrodinger, Prigogine, Gell-Mann, Eigen, Turing, Smale, Hodgkin-Huxley, et al., had spent much of the later part of their life, toiling in vain to solve them. For example, in 1781, Galvani was dumfounded when the exposed nerve of a frog leg he was dissecting suddenly twitched when hit by a spark from a nearby Leyden jar. Galvani has spent countless sleepless nights trying to explain the phenomenon by seeking for an "extremely mobile principle", which we now know is the local activity principle. Fast forward 100 years later, Boltzmann was struggling to explain how life could emerge, and concluded, in a qualitative way, that the "existence of animate beings is a struggle for entropy". About 60 years later, Schrodinger continued Boltzmann's incomplete quest with his discovery that what Boltzmann needs is "negative entropy" -- an intuitive but fuzzy concept that we can now identify as the existence of an edge of chaos domain.  Prigogine continued Schrodinger's thermodynamic approach by proposing his "Instability of the homogeneous principle", which turns out to be astability condition that was misleading, and had actually impeded progress, when interpreted from the fine lens of the Edge of Chaos Corollary -- the pearl of the local activity principle. Gell-Mann came very close to the target when he proposed the holy grail to be the system's inherent " ability to amplify small signal fluctuations". He failed because he did not realize that the relevant physical variable to be endowed this small-signal gain property must be the local energy about an equilibrium state. To illustrate the universality of the local activity principle, I will touch base with Alan Turing, father of computer science and AI, in particular his contribution on the "Chemical basis of Morphogenesis". He was perplexed on how an embryo in its spherical blastula stage could break symmetry, without external forcing, to deform and grow into a non-symmetrical embryo. With great intuition, Turing devised a toy example of 2 identical cells each harboring 2 chemicals with concentrations X and Y, respectively. When uncoupled, the 2 chemicals in the 2 identical cells have identical concentrations, as expected. By applying trial-and-error, and intuition, Turing selected 2 different diffusion coefficients D(X) and D(Y) to couple the 2 identical cells. Although diffusion is a dissipative process, Turing's ingenious insight suggested the concentrations of the 2 identical cells would change, and the symmetry of the resulting reaction-diffusion equations was broken!  Turing's choice of the 2 diffusion coefficients that led to symmetry breaking was based on his intuitive guess, because Turing did not have an algorithm for choosing the 2 dissipative coefficients that would guarantee symmetry breaking. This is the occasion where the edge of chaos criterion comes in handy, for it provides a simple "symmetry breaking" algorithm for calculating the 2 diffusion coefficients. I will present an example in this talk by choosing 2 identical first-order nano-scale NbO2  memristor circuits and couple them via a positive resistance, whose value is calculated via the Edge-of-Chaos Criterion such that the two identical first-order memristor circuits in the resulting reaction-diffusion equations have different equilibrium states. Indeed, I will follow Stephen Smale to couple two identical 2nd-order nano-scale NbO2  memristor circuits via 2 positive resistors calculated to be on their respective edge-of-chaos domain, and demonstrate to the audience that the symmetry of the resulting 4th-order reaction-diffusion equations is broken, resulting in a 4-dimensional oscillator.

Sensing technologies for Machine Vision- Device to System

Speaker Biosketch

Ahmad Bahai, Ph.D, is a senior vice president and chief technology officer (CTO) of Texas Instruments responsible for guiding break-through innovation, corporate research and Kilby Labs.

Dr.  Bahai is an Adjunct professor at Stanford University, and IEEE Fellow. He was a professor in residence at UC Berkeley from 2001-2010. Throughout his career, Dr. Bahai has held a number of leadership roles including director of research labs and chief technology officer of National Semiconductor, technical manager of a research group at Bell Laboratories and founder of Algorex, a communication and acoustic IC and system company, which was acquired by National Semiconductor. He holds a Master of Science in Electrical Engineering from Imperial College, University of London and a doctorate degree in Electrical Engineering  from University of California, Berkeley. Ahmad Bahai, Ph.D, is a senior vice president and chief technology officer (CTO) of Texas Instruments responsible for guiding break-through innovation, corporate research and Kilby Labs.

Abstract

Dr. Ahmad Bahai, Chief Technology Officer and Senior Vice President at Texas Instruments will share his perspective and vision on new horizons for machine vision.  This talk will provide an overview of the rapidly evolving sensing technologies including radar, lidar and ultrasonic sensing and the research and development opportunities at device, packaging and system/algorithm. Technologies such as low cost deep submicron CMOS, Si-Ge, and other compound materials are promising from different performance & figure of merit criteria which in many cases demands a hybrid integration as a SIP. This talk will cover some of the current and upcoming technologies and trade-offs.

University-Industry Collaboration: Perspectives on Challenges and Opportunities

Speaker Biosketch

Terri Fiez is the Vice Chancellor for Research and Innovation at University of Colorado Boulder (CU Boulder).  She oversees the >$600 million research enterprise, the technology transfer organization and the cross-campus Innovation and Entrepreneurship Initiative.  Dr. Fiez is passionate about infusing innovative thinking and an entrepreneurial approach while building connections across the region to enhance U.S. competitiveness.  The Association of Public and Land Grant University (APLU) recently awarded CU Boulder the 2021 Innovation, Entrepreneurship and Technology-Based Economic Development Award.  CU Boulder is also a co-lead for a new NSF I-Corps Hub in the Western Region with USC and UCLA.

Prior to 2015, Dr. Fiez was Director of the School of Electrical Engineering and Computer Science at Oregon State University.   From 1999-2015 she oversaw tremendous growth in new innovation education programs and in the research and Ph.D. programs.  In 2008 she co-founded a solar electronics startup company (Azuray Technologies), raised venture capital funding and took a leave of absence to build the company.  She was bitten by the startup bug and seeks to seed entrepreneurial experiences for students and faculty. 

Dr. Fiez received her B.S. and M.S. degrees in Electrical Engineering at the University of Idaho, Moscow, in 1984 and 1985, respectively.  She received her Ph.D. degree from Oregon State University in Electrical and Computer Engineering in 1990.  Dr. Fiez’s research focuses on analog and mixed-signal integrated circuits and innovative education experiences. She holds 4 U.S. patents, 1 foreign patent and 1 copyright licensed. She has published over 150 conference and journal papers, advised 85 M.S. and Ph.D. students and has served as Distinguished Lecturer for Institute of Electrical and Electronics Engineering (IEEE) Solid-State Circuits and IEEE Circuits and Systems Societies, and as Associate Editor of IEEE TCASII and Guest Editor for IEEE Journal of Solid-State Circuits.

She received the NSF Young Investigator Award early in her career.  She used this early career award to build connections with industry in support of research and teaching.  Through her commitment to unique education opportunities, she was awarded the following international awards the 2006 IEEE Educational Activities Board’s Innovative Education Award, the 2016 IEEE Undergraduate Teaching Award and the 2017 IEEE Education Society Harriett B. Rigas Award.  Dr. Fiez is a Fellow of IEEE and of the National Academy of Inventors.

Abstract

University faculty and students often seek to collaborate with companies in research, internships, and bringing real-world perspectives to the classroom.  This exposure can be very beneficial to students and also enhances the relevance of university research.  Companies are motivated to engage with universities so that they have a prepared workforce and can tap into the intellectual creativity of the research.  On the surface, it looks like a match made in heaven.  In this talk, conversations with a few dozen company engineering leaders and university faculty from around the world will illuminate challenges faced while recognizing the tremendous synergies and opportunities.  The goal of this talk is to plant seeds for how universities and companies around the world can chart a path that creates the win-win.

Extracting More Compute from Each Transistor

Speaker Biosketch

Samuel Naffziger is AMD senior vice president, Corporate Fellow, and Product Technology Architect. Naffziger works across the company to optimize product technology choices and deployment with a continued focus on driving best practice power/performance/area methodology to maximize product competitiveness, efficiency, and cost. Naffziger has been the lead innovator behind many of AMD’s low-power features and chiplet architecture. He has over 32 years of industry experience with a background in microprocessors and circuit design at Hewlett Packard, Intel and AMD. Naffziger received a Bachelor of Science degree in Electrical Engineering from the California Institute of Technology (CalTech) and a Master of Science from Stanford. Naffziger holds more than 130 U.S. patents in the field and authored dozens of publications and presentations on processors, architecture and power management. He is an IEEE Fellow.

Abstract

For decades, Moore's Law has delivered the ability to integrate an exponentially increasing number of devices in the same silicon area at a roughly constant cost. This has enabled tremendous levels of integration, where the capabilities of computer systems that previously occupied entire rooms can now fit on a single integrated circuit.  It's well documented that Moore's Law is slowing on all its dimensions and corollaries. While improvements in density scaling continue at a reduced pace, the industry is simultaneously observing increases in manufacturing costs that are threatening to eliminate the cost per transistor generational improvement that has been fundamental to the value proposition of semiconductor compute products.  In response to this existential threat to the industry, the solution is, as it always has been, to innovate.

This talk will focus on several compelling vectors of innovation that will enable our industry to thrive within the limits of physics and silicon scaling. The first is around extraction of more value from each manufactured transistor through modular chiplet architectures that enable application-specific choice of process technology and market-specific personalization of products. The second opportunity extracts more performance from each yielded transistor through targeted use of special purpose accelerators that provide dramatic efficiency gains over general purpose compute. Finally, there is the growth in applications of AI to all aspects of computing, making use of reduced precision accelerators that add non-linear value to compute capabilities and user experience in all key market segments. Putting all these together with the right software solutions will provide synergistic improvements in compute capability that can enable a thriving industry for the foreseeable future.