User:104gli/Kurt Keutzer braindump

Kurt Keutzer is an American professor, computer scientist and executive.

Articles to parse (Temporary)

 * Predicting the rise of SOCs and other stuff:
 * Interview that explains Keutzer's path from EDA to Parallel Computing:
 * ParLab summaries:
 * "As Deep Neural Networks have proved successful at providing state-of-the-art solutions to all of these problems, I am now focusing my entire research efforts on the problems of providing state-of-the-art solutions to the training and deployment of Deep Neural Networks."

Facts to organize (Temporary)

 * "Keutzer, Kurt William was born on November 9, 1955 in Indianapolis, Indiana, United States. Son of William Dale and Mary Helen Keutzer."
 * As of 2010: 6 books and over 200 refereed articles
 * Fellow of IEEE
 * "Kurt has been an investor and advisor to thirteen startups and an advisor to seven more."
 * John Shalf: I think Kurt Keutzer, one of the lead authors on the report, sums this up best when he says "This shift toward increasing parallelism is not a triumphant stride forward based on breakthroughs in novel software and architectures for parallelism; instead, this plunge into parallelism is actually a retreat from even greater challenges that thwart efficient silicon implementation of traditional uniprocessor architectures."

Some things I'd like to cover

 * Grew up in Indiana
 * Early interest in Math, Buddhism, and Tibetan Studies
 * Research story arc
 * Math
 * Math experts went into Electronic Design Automation (EDA) - intern at Bell Labs
 * Joined Bell Labs
 * Synposys (recruited by Aart de Geus)
 * rise to CTO
 * Berkeley (recruited by Alberto Sangiovanni-Vincentelli)
 * From EDA to Parallel Computing. ParLab.
 * In [year], [David Patterson], Kurt Keutzer, and N other professors co-founded the Parallel Computing Laboratory (ParLab) at UC Berkeley.
 * From Parallel Computing to Machine Learning to Deep Learning
 * BVLC, BAIR, and BDD
 * SqueezeNet and friends
 * Accelerating DNN Training: (FireCaffe), (LARS  )
 * TinyML
 * DeepScale
 * Bibliography -- cover all published books, use Paul Halmos' wiki page for inspiration
 * Entrepreneurship
 * Invested in 13 companies
 * Co-founded DeepScale
 * Career advice
 * "Entrepreneurship is the only career for the 21st century."
 * Tibetan Studies
 * Research at the intersection of Tibetan Studies and Machine Learning -- e.g. Namsel Tibetan OCR project.

Futurology

 * When I was working at Synopsys, a marketing director took the trouble to come to my office late on a Friday afternoon to express his concern that our company's focus on system-on-chip (SoC) was missing the market reality. "Only 3 percent of chips currently would be described as SoCs!" he lamented. He was concerned that I and my co-conspirators were going to drag the company down the tubes with an SoC focus.
 * Predicting that parallelism will be important in ML [REF]
 * Predicting that we can bring down the costs of DL training and inference [REF]
 * Predicting that memory and I/O, not compute, will be the bottleneck in DL training and inference [REF]

Books by Keutzer

 * 1988. Dwight Hill, Don Shugard, John Fishburn, and Kurt Keutzer. Algorithms and Techniques for VLSI Layout Synthesis. Springer.
 * 1994. Srinivas Devadas, Abhijit Ghosh, and Kurt Keutzer. Logic Synthesis. McGraw-Hill.
 * 2002. David Chinnery and Kurt Keutzer. Closing the Gap Between ASIC & Custom: Tools and Techniques for High-Performance ASIC Design. Springer. (2nd edition appeared in 2007. )
 * 2004. Pinhong Chen, Desmond A. Kirkpatrick, and Kurt Keutzer. Static Crosstalk-Noise Analysis: For Deep Sub-Micron Digital Designs. Springer.
 * 2005. Matthias Gries and Kurt Keutzer. Building ASIPs: The Mescal Methodology. Springer.

Honors and awards

 * IEEE Fellow (1996)
 * "Top 10 Most Cited Author" and "Author of a Top 10 Most Cited Paper" at 50th Design Automation Conference (YEAR)
 * (Various best-paper awards)