PeopleofSuperintelligence

Researcher

Kurt Keutzer

Kurt Keutzer is a computer scientist whose research makes machine-learning systems more computationally efficient. He earned a B.S. in mathematics at Maharishi University and a PhD in computer science at Indiana University, worked at Bell Labs and Synopsys, was elected an IEEE Fellow in 1996, and joined Berkeley in 1998. FireCaffe used GPU clusters to train neural networks, while SqueezeNet matched AlexNet-level ImageNet accuracy with 50 times fewer parameters; Keutzer coauthored both and co-founded DeepScale. His 1987 Dagon paper received the DAC's Most Influential Paper Award in 2023, and he coauthored a study of multi-agent LLM failures in 2025. As of October 2026, he is Professor in the Graduate School at Berkeley, researching efficient large language models and agents.

Born
1955
Badges

Years link to their sources.

Timeline

  1. 1978–1984

    Degrees and Bell Labs

    He earns a B.S. in mathematics at Maharishi University in 1978 and a PhD in computer science at Indiana University in 1984. After earning his PhD, he joins AT&T Bell Laboratories' research division.

  2. 1991

    Joining Synopsys

    He joins Synopsys in 1991 and becomes its Chief Technical Officer and Senior Vice-President of Research.

  3. 1996

    An IEEE Fellowship

    He is elected a Fellow of the IEEE in 1996.

  4. 1998

    Professor at Berkeley

    In 1998, he joins UC Berkeley's EECS faculty as a professor.

  5. 2015–2018

    FireCaffe, SqueezeNet, and DeepScale

    In 2015, he coauthors FireCaffe, which trains GoogLeNet and Network-in-Network on ImageNet with 47x and 39x speedups on 128 GPUs. In 2016, he coauthors SqueezeNet, which reaches AlexNet-level accuracy on ImageNet with 50 times fewer parameters. He co-founds DeepScale with Forrest Iandola in 2017 and serves as its Chief Strategy Officer in 2018.

  6. 2023

    A DAC award for Dagon

    In 2023, his 1987 paper “Dagon: technology binding and local optimization by DAG matching” wins the DAC's Most Influential Paper Award for the 1980s.

  7. March 2025

    Studying multi-agent failures

    In March 2025, he coauthors “Why Do Multi-Agent LLM Systems Fail?” The study identifies 14 unique failure modes, grouped as system design issues, inter-agent misalignment, and task verification.

  8. As of October 2026

    Professor in the Graduate School

    As of October 2026, he is a Professor in the Graduate School at UC Berkeley. His current research focuses on making large language models and agents more efficient.

Career

  1. Since 1998University of California, BerkeleyProfessor in the Graduate School
  2. From 2017DeepScaleCo-founder, Chief Strategy Officer
  3. From 1991SynopsysChief Technical Officer and Senior Vice-President of Research
  4. From 1984Bell Labs

Education

  1. 1984Indiana UniversityPhD
  2. 1978Maharishi UniversityB.S.

From the biography.

  1. 11% of AI doctorate holders from 2010 to 2016 founded or led a company or lab by 10 years after graduationIncludes Kurt Keutzer. We estimate how often AI doctorate holders founded or led a company or lab by 10 years after graduation, how waiting times compare between founders and top jobs, and which universities trained the most.
  2. 23% of people shaping AI who left a university for a company later returned to a universityIncludes Kurt Keutzer. We estimate that 468 of the 2,011 people shaping AI who left a university for a company (23%) later returned to a university after a median of 5 years, and 163 of 829 (20%) made the reverse round trip.
  3. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes University of California, Berkeley. We estimate where the people shaping AI studied and where their careers led. Massachusetts Institute of Technology leads with 467 former students, including 291 recorded doctorates.
  4. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes University of California, Berkeley. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Kurt Keutzer yet. Send us a free-licensed photo.

  1. 03.11.98 - UC Berkeley New Faculty, 1997-1998 newsarchive.berkeley.edu. Checked 20 October 2026.
  2. Kurt Keutzer people.eecs.berkeley.edu. Checked 20 October 2026.
  3. Kurt Keutzer - Wikipedia en.wikipedia.org. Checked 20 October 2026.
  4. FireCaffe: near-linear acceleration of deep neural network training on compute clusters arxiv.org. Checked 20 October 2026.
  5. SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size arxiv.org. Checked 20 October 2026.
  6. Kurt Keutzer people.eecs.berkeley.edu. Checked 20 October 2026.
  7. Kurt Keutzer receives DAC Most Influential Paper Award eecs.berkeley.edu. Checked 20 October 2026.
  8. Why Do Multi-Agent LLM Systems Fail? arxiv.org. Checked 20 October 2026.
  9. Kurt Keutzer | EECS at UC Berkeley www2.eecs.berkeley.edu. Checked 20 October 2026.
  10. Kurt Keutzer - International Computer Science Institute (ICSI) icsi.berkeley.edu. Checked 20 October 2026.