Challenges in Analog/Mixed-Signal Design Automation
University of Minnesota
The traditional view of electronic design automation has intensively focused on the design, synthesis, and layout of digital circuits. This perspective has been reinforced by the trends from Moore's law, which have seen digital system complexities grow exponentially, prompting an acute need for efficient design tools and flows. In contrast, analog design has remained largely focused on the expert designer. This world view is now changing, for several reasons. First, several tasks in analog design are now at a point where they can be realistically automated, notably tasks related to layout automation. In advanced finFET technologies, the reduction in the degrees of freedom due to restricted design rules actually makes layout automation easier.
Second, the clear distinction between analog and digital designs has blurred, with modern designs seeing a great deal of digital-like circuitry that assists in implementing analog functionalities. For these structures, established techniques from digital system design can carry over to enable design automation. Third, the complexity of the mixed-signal design space makes it difficult for designers to fully comprehend and compensate for the impact of phenomena such as process variations and device aging. Especially under stringent design specifications, these complexities create openings for design automation tools that can complement the knowledge of the expert designer. Thus, analog and mixed-signal design, which has long been the bastion of the expert designer, is projected to be the new frontier in design automation. This talk will present a brief history of prior efforts and will overview the set of opportunities and challenges in this emerging field.
Prof. Sachin S. Sapatnekar
University of Minnesota
Prof. Sachin Sapatnekar received the Ph.D. degree from the University of Illinois at Urbana-Champaign in 1992, after which he joined the faculty at Iowa State University. Since 1997, he has been teaching at the University of Minnesota, where he is a Distinguished McKnight University Professor and the Henle Chair in Electrical and Computer Engineering. His research is related to developing CAD techniques for the analysis and optimization of circuit performance, currently focused on both CMOS circuits and spintronics technologies. He has served as Editor-in-Chief of the IEEE Transactions on CAD and General Chair for the ACM/IEEE Design Automation Conference (DAC). He is a recipient of several conference Best Paper Awards, ten-year retrospective Best Paper Awards, the Semiconductor Research Corporation's Technical Excellence Award, and the Semiconductor Industry Association University Research Award, and a Fulbright award. He is a Fellow of the IEEE and the ACM.
Magnetic and inductive position sensors for industrial and automotive application. From silicon design to system design
Intelligent magnetic and inductive position sensor solutions gain more and more market share compared to optical, resolver and other technologies. High level CMOS integration on digital, analog and high voltage periphery elements support an accurate, cost-effective and reliable sensor solution. This speech will give an overview about the implementation path of a monolithic magnetic sensor IC together with system level interactions like magnetics and mechanics.
This presentation will cover industrial and automotive use cases of a position sensor IC and will highlight the multidisciplinary interaction in sensor systems. This talk will compare different position sensor technologies and will highlight important functions of integrated circuits. Several new market trends in robotics, automotive e.g. autonomous driving and electrification will further drive IC developments and further shrink of process technology nodes. Time to market will further shorten development times in engineering areas in general.
Marcel joined ams AG as IC design engineer for magnetic position Sensors 2000. He is currently Head of Product- and Application Management in the ams AG business line Position Sensors. Marcel graduated at the Carinthian Tech Institute located in Villach and received a master degree in electrical engineering. He is leading the application support team for position sensors in Austria and is covering automotive, industrial and consumer market Marcel holds several patents for magnetic position sensors and published several technical articles in this domain.
Accelerating Mixed-Signal Design Verification: Turn a SPICE netlist into a SystemVerilog Model
University of Washington
Design verification is a bottleneck on modern SoC design. Very often, modern SoCs contain both digital blocks and analog blocks. While the metric-driven verification methodology exists for digital verification, analog and mixed-signal design verification relies heavily on SPICE simulation and manual modeling, a process known to be time consuming and error prone. In this talk, we will introduce a new technology that can turn automatically a SPICE netlist into a SystemVerilog model, and thus allows metric-driven digital verification methodologies and tools to be used for analog and mixed-signal design verification. This breakthrough is based on a new theory of signal abstraction developed under a recent DARPA sponsored research program.
We will show that how signal-driven abstraction allows various circuit analysis techniques developed in the past several decades including symbolic analysis, Laplace transform, pole/zero extraction and fractional expansion, event-driven analog modeling, interval mathematics, modified nodal analysis, regression, wreal and real number modeling, language compilation, analog assertion, can all be integrated in this unified framework, and to be used in a methodology transparent to designers. Practical examples from 28Gbps SerDes design will be used to illustrate the methodology. In particular, this talk will show the design verification of 28-Gb/s serial transceiver link adaptation and equalization, which were not feasible previously. This research has been supported by the US DARPA IRIS program.
Prof. Richard Shi
University of Washington
Prof. Richard Shi is currently a Professor in Electrical Engineering at the University of Washington, Seattle WA, where he joined in 1998. He received a prestigious Doctoral Prize from the Natural Science and Engineering Research Council (NSERC) of Canada and a Governor-General's Silver Medal in 1995 for his PhD Dissertation in computer science. His current research interest includes energy-efficient circuit and system design for sensing, computing, learning and communication. Since 2006, he has directed several DoD-sponsored projects in the area of ADC, PLL, SerDes and LDPC design. Previously, he worked in the area of computer-aided design of mixed-signal integrated circuits, in which for his contribution he was elevated to a Fellow of IEEE in 2005. He received several awards for his research including Donald O. Pederson Best IEEE Transactions on CAD Paper Award, Best Paper Awards from the IEEE/ACM Design Automation Conference, the IEEE VLSI Test Symposium, and the SRC Technical Conference, and an NSF CAREER Award. He has served ten years as an Associate Editor of the IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems twice as an Associate Editor, once as a Guest Editor, of the IEEE Transactions on Circuits and Systems-II: Analog and Digital Signal Processing, as well as Transactions Briefs. He is a key contributor to the IEEE 1076.1 standard on mixed-signal circuit description and simulation language. He co-founded three startups in EDA and one in chip design, which have been acquired. He is a founder of Orora Design Technologies, Inc., a pioneer in automating mixed-signal design verification.
Innovating Vision Beyond the Human Eye - Navigating the Expanding CMOS Image Sensor Design Space
Today’s image sensor solutions for demanding applications like automotive driver assist image sensing, high speed machine vision, and low power battery operated cameras started from relatively modest solutions for pixels and analog readout design. These initial solution options have evolved along different paths. Now the designer must understand the overall image system requirements to decide where and what to process in the pixel domain, analog circuit domain, and digital processing domain.
These decisions rely on understanding the fundamentals of imager photon capture, noise components, analog/digital signal processing topology efficiency, and now the impact of 3D wafer stacking on sensor architecture solutions. This talk will review some of the fundamentals of image sensor technology, design, how it evolved to its current state, and the trajectory for what is possibly next.
Roger Panicacci is Vice President of Imager Design Platform Technology at On Semiconductor where he focuses on circuits, sensor architecture, and system design for next generation image sensor products. He started work on CMOS active pixel image sensor design at the Jet Propulsion Laboratory in the 1990s during the early development of the technology and was one of founding members of Photobit Corporation that commercialized CMOS active pixel technology. He stayed with the same team as it became part of Micron Technology, then Aptina, and now On Semiconductor. He holds a BS EECS from the University of California, Berkeley, and MS EE from the University of California, Santa Barbara.