PeopleofSuperintelligence

Lab leader

Bryan Catanzaro

Bryan Catanzaro is vice president of Applied Deep Learning Research at NVIDIA, where he heads the Nemotron team that builds open foundation models. He interned at Intel in 2001, then started a PhD at Berkeley in 2005, where a search for demanding parallel applications led him to machine learning. After graduating in 2011, he worked at NVIDIA Research, and in 2014 he co-authored the paper that introduced cuDNN, a library of efficient deep-learning primitives. He co-authored the 2019 Megatron-LM paper on training transformer models with billions of parameters. In December 2025 he gave the opening address at NVIDIA's Nemotron Summit at NeurIPS.

Timeline

  1. 2001

    An internship at Intel

    He interns at Intel in 2001, building circuits for future CPUs that were supposed to run at 10 GHz.

  2. 2005

    A PhD at Berkeley

    He starts a PhD at Berkeley in 2005, looking for parallel applications with heavy computing demands, and that search leads him to machine learning.

  3. 2009

    An NVIDIA Graduate Fellowship

    In 2009 NVIDIA names him one of its Graduate Fellows while he studies at the University of California, Berkeley.

  4. 2011–2014

    NVIDIA Research and cuDNN

    After graduating from Berkeley in 2011, he works at NVIDIA Research. In 2014 he co-authors the arXiv paper that introduces cuDNN, a library of efficient deep-learning primitives.

  5. By 2016

    Speech systems at Baidu

    Before returning to NVIDIA in 2016, he works at Baidu with Adam Coates and Andrew Ng on end-to-end speech recognition systems.

  6. 2016–2019

    Applied Deep Learning Research lab

    The lab he heads builds prototype applications for NVIDIA. In 2019 he co-authors the Megatron-LM paper, which describes training transformer models with billions of parameters.

  7. 2025

    Open Nemotron models

    In September 2025 he writes an NVIDIA blog post on the Nemotron models, datasets and techniques. In December 2025 he gives the opening address at the Nemotron Summit at NeurIPS.

  8. As of October 2026

    Heading the Nemotron team

    He is vice president of Applied Deep Learning Research at NVIDIA, heading the Nemotron team, which released Nemotron 3 Nano in December 2025.

Sources and credits

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

  1. Working AI: With VP of Applied Deep Learning Research Bryan C. deeplearning.ai. Checked 14 October 2026.
  2. 2009 Grad Fellows | Research research.nvidia.com. Checked 14 October 2026.
  3. [1410.0759] cuDNN: Efficient Primitives for Deep Learning arxiv.org. Checked 14 October 2026.
  4. Bryan Catanzaro | TWIML - The Voice of Machine Learning & AI twimlai.com. Checked 14 October 2026.
  5. [1909.08053] Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism arxiv.org. Checked 14 October 2026.
  6. Nemotron Models, Datasets and Techniques Fuel AI Development | NVIDIA Blog blogs.nvidia.com. Checked 14 October 2026.
  7. NVIDIA Advances Open Model Development for Digital and Physical AI | NVIDIA Blog blogs.nvidia.com. Checked 14 October 2026.
  8. Meet Bryan Catanzaro, speaker at HumanX | San Francisco | April 6-9, 2026 humanx.co. Checked 14 October 2026.
  9. Why Nvidia builds open models with Bryan Catanzaro interconnects.ai. Checked 14 October 2026.