PeopleofSuperintelligence

Researcher

Russell J. Hewett

Russell J. Hewett is a computational scientist who works on training deep learning models at scale, from model-parallel research at Virginia Tech to language models at Microsoft and NVIDIA. After a computer science PhD at the University of Illinois at Urbana-Champaign in 2011, he spent 2011 to 2014 as a postdoctoral associate in applied mathematics at MIT. As an assistant professor of mathematics at Virginia Tech, he co-developed the DistDL package for model-parallel deep learning in PyTorch and won a Department of Energy Early Career Research Program award in 2021. By 2023 he had joined Microsoft, where he coauthored the Phi-4 Technical Report in December 2024, and he is among NVIDIA's authors on the NVFP4 pretraining paper and the Nemotron 3 reports, including Nemotron 3 Ultra in June 2026.

Now at
NVIDIA
Works
77
Citations
1,414
h-index
14
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2005–2011

    Degrees from Virginia Tech and Illinois

    He earns a BS with honors in computer science at Virginia Tech in 2005. In 2011, he completes a PhD in computer science at the University of Illinois at Urbana-Champaign.

  2. 2011–2014

    Postdoc in MIT mathematics

    From 2011 to 2014, he is a postdoctoral associate in applied mathematics at the Massachusetts Institute of Technology, where he belongs to the Imaging and Computing Group in the mathematics department.

  3. 2020

    Model parallelism and DistDL

    In June 2020, he and Thomas J. Grady II post "A Linear Algebraic Approach to Model Parallelism in Deep Learning" on arXiv. The paper treats parallel data movement operations such as broadcast, sum-reduce and halo exchange as linear operators with hand-built adjoints, and trains a distributed network with DistDL. DistDL is their Python package for model-parallel deep learning in PyTorch.

  4. 2021

    A DOE Early Career award

    He wins a Department of Energy Early Career Research Program award in 2021, through the Advanced Scientific Computing Research program, for a project on domain-decomposition-induced parallelism for scientific deep learning at extreme scale. He holds the post of assistant professor of mathematics at Virginia Tech, with an affiliation in Computational Modeling and Data Analytics, after working as a research scientist and project manager in computational geoscience in Houston.

  5. 2023

    From Virginia Tech to Microsoft

    He coauthors "Model-Parallel Fourier Neural Operators as Learned Surrogates for Large-Scale Parametric PDEs", which predicts PDE solutions of over 2.6 billion variables on Perlmutter with up to 512 A100 GPUs and trains a distributed Fourier neural operator on the Azure cloud. The research is performed at Virginia Tech, and by the February 2023 version his affiliation is Microsoft.

  6. December 2024

    The Phi-4 Technical Report

    He is one of the authors of Microsoft Research's Phi-4 Technical Report, published in December 2024. The report presents phi-4, a 14-billion parameter language model trained with a recipe centered on data quality that uses synthetic data throughout training.

  7. September 2025–December 2025

    NVFP4 and Nemotron 3 Nano

    He is among the NVIDIA authors of "Pretraining Large Language Models with NVFP4" in September 2025, which trains a 12-billion-parameter model on 10 trillion tokens in 4-bit floating-point precision. In December 2025, he is an author of NVIDIA's Nemotron 3 Nano report, which describes a 30B-A3B Mixture-of-Experts hybrid Mamba-Transformer model pretrained on 25 trillion text tokens.

  8. As of October 2026

    Nemotron 3 Ultra at NVIDIA

    He is among NVIDIA's authors on the Nemotron 3 Ultra report submitted in June 2026. The report introduces a Mixture-of-Experts Hybrid Mamba-Attention model with 550 billion total and 55 billion active parameters, pre-trained on 20 trillion text tokens and extended to a 1M-token context length.

Career

  1. NowNVIDIA
  2. Until 2024Microsoft Research
  3. 2015–2020Total
  4. 2011–2014Massachusetts Institute of TechnologyPostdoctoral Associate
  5. Virginia TechAssistant Professor of Mathematics

Education

  1. 2011University of Illinois Urbana-ChampaignPhD
  2. 2005Virginia TechBS with honors in computer science

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. The SunPy Project: Open Source Development and Status of the Version 1.0 Core Package2020 · The Astrophysical Journal · 495 citations
  2. SunPy—Python for solar physics2015 · Computational Science & Discovery · 120 citations
  3. A Robust Null Space Method for Linear Equality Constrained State Estimation2010 · IEEE Transactions on Signal Processing · 86 citations
  4. Multifractal Properties of Evolving Active Regions2007 · Solar Physics · 67 citations
  5. Dynamic Three-Dimensional Tomography of the Solar Corona2010 · Solar Physics · 55 citations

From OpenAlex, as of 21 October 2026.

  1. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Russell J. Hewett. We estimate that alumni of 17 major AI labs went on to found or lead 350 other organizations, including 226 they founded, and we trace 472 career moves directly between 23 labs since 2010.
  2. Google has the highest combined h-index among AI organizationsIncludes NVIDIA. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  3. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes NVIDIA. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.
  4. Universities (31%) and startups and other companies (25%) employ the largest shares of 2020–2026 entrantsIncludes NVIDIA. Among 819 people whose first recorded role started in 2020–2026 and who have a known current organization, 60% work in commercial sectors (25% at startups and other companies, 23% at frontier labs and 13% at big tech, with 9 in two), and 31% work at universities, compared with 66% for earlier entrants.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Russell J. Hewett yet. Send us a free-licensed photo.

  1. rjh.io | russell j. hewett rjh.io. Checked 20 October 2026.
  2. [2006.03108] A Linear Algebraic Approach to Model Parallelism in Deep Learning arxiv.org. Checked 20 October 2026.
  3. Distributed Deep Learning | rjh.io | russell j. hewett rjh.io. Checked 20 October 2026.
  4. Hewett awarded a DOE Early Career Research Program Award | Department of Mathematics | Virginia Tech math.vt.edu. Checked 20 October 2026.
  5. Model-Parallel Fourier Neural Operators as Learned Surrogates for Large-Scale Parametric PDEs arxiv.org. Checked 20 October 2026.
  6. Phi-4 Technical Report - Microsoft Research microsoft.com. Checked 20 October 2026.
  7. [2412.08905] Phi-4 Technical Report arxiv.org. Checked 20 October 2026.
  8. [2509.25149] Pretraining Large Language Models with NVFP4 arxiv.org. Checked 20 October 2026.
  9. [2512.20848] Nemotron 3 Nano: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning arxiv.org. Checked 20 October 2026.
  10. [2606.15007] Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning arxiv.org. Checked 20 October 2026.