Lynn Conway 1938-2024
When I interviewed Rob Rutenbar just after he was announced as the honoree of the 2017 Kaufman Award, he said “I was a Mead and Conway baby”. As it happens, I was a Mead and Conway baby too. If you are too young to have experienced this, you might not even know what I am talking about. In 1979 the book Introduction to VLSI Systems was published. Nobody ever called it that though, it was always just Mead & Conway. See my old post The Book That Changed Everything.
I told some the story of how I came to become a Mead and Conway Baby there:
Mead and Conway separated design from manufacturing by creating simplified design rules for layout, and a simplified timing model suitable for digital design. No longer was it necessary to understand every nuance of the fabrication process, no longer was it necessary to consider every transistor as an analog device. The most important aspect of this is that it meant that computer scientists could design digital chips because they no longer needed deep electrical engineering knowledge.
I was at Edinburgh University at the time, finishing up my Ph.D. (in computer science, not electrical engineering). John Gray ran Carver Mead’s silicon structures project at Caltech while on sabbatical from Edinburgh, and Irene Buchanan had gone there too. She returned carrying galley proofs of the yet-to-be-published Mead and Conway book. She ran a course based on this, one of the first on IC design in the UK, I presume, heavily attended not just by the final-year students who were meant to be on the course, but by many of us faculty as well.
Since the book was published in 1979, I guess this must have been in 1978.
As I type this, probably a few days before you will read it, I am here at the Design Automation Conference in San Francisco. In the section covering honorees for the Marie Pistilli Award (Synopsys’ Sashi Obilisetty) and the DAC Under-40 Innovators Award, there was a tribute to Lynn Conway, who passed away earlier this month. We had a moment of silence for her.
It is hard to overstate how much the semiconductor industry owes to Lynn Conway and her co-author Carver Mead. Prior to this book, semiconductor design was an arcane art practised only inside semiconductor companies. It was not taught at University since nobody teaching knew anything about it. That all changed with the publication of Introduction to VLSI Systems and some of the other infrastructure that sprung up, such as summer schools and free multi-project wafer manufacturing (today we would call it a foundry but we did not back then). Almost overnight a generation of computer scientists suddenly learnt how to design chips. If you know anything about computer scientists, you can guess correctly that these people immediately started to automate the process, creating what today we call EDA, for Electronic Design Automation. I, for example, wrote VLSI Technology’s circuit extractor that read in layout and wrote out a netlist of transistors and capacitors ready for simulation and other verification. Resistance and lateral capacitance were still low enough to ignore, unlike today. Without Mead and Conway I would never have considered joining VLSI Technology, and probably even the company would not have existed (in anything like the form that it did, anyway).
Immediately after that session was the opening keynote by Jim Keller of Tenstorrent. It wasn’t the focus of his keynote, but he talked in passing about some of the challenges of creating the first out-of-order execution processors. It turns out that Lynn Conway had a hand in that too. In the first part of her life, Lynn was male and worked at IBM, publishing (with co-authors) the earliest papers about scheduling multiple instructions in parallel, something that all high-end microprocessors do today but back then even mainframes did not. She uses the name Robert Sanders for that phase of her life, but that was not her actual name. When IBM discovered that she planned to transition, she was fired (IBM apologized for this in 2020). She did transition and then started her life again as a woman. As part of that life, she ended up at Xerox Palo Alto Research Center (PARC) working on VLSI design. And writing the book that changed everything.
You can find more about Lynn Conway’s life in my Semiwiki post from 2013 Lynn Conway’s Story.
You can read the IEEE obituary Honoring the Legacy of Chip Design Innovator Lynn Conway.