PeopleofSuperintelligence

Builder

Pearu Peterson

Also known as P Peterson

Pearu Peterson is a scientific-software engineer whose work spans NumPy, SciPy and sparse tensor support in PyTorch. After a 2001 PhD on soliton theory, he created the f2py Fortran-to-Python tool, which ships as part of NumPy, and in 2001 he merged code with Travis Oliphant and Eric Jones to form SciPy. By 2011 he was a research scientist at Simula Research Laboratory in Oslo, and in 2020 he co-authored the SciPy 1.0 and NumPy papers. Between 2021 and 2022 he wrote about the CSR layout and other sparse tensor work for PyTorch. He is a Staff Software Engineer at OpenTeams and is listed among NumPy's maintainers.

Now at
OpenTeams
Born
1971
Links
Website

Years link to their sources.

Timeline

  1. 2001

    A PhD on solitons

    He receives a PhD in 2001 in soliton theory, covering multi-soliton interaction patterns and inverse problems.

  2. 2001

    Releasing f2py

    In December 2001, he announces the third public release of f2py, a command line tool that scans Fortran 77/90/95 code and generates a Python C/API module. A paper on the tool, written with Joaquim R. R. A. Martins and Juan J. Alonso, appears in the proceedings of the 9th International Python Conference in 2001. Originally created by him, F2PY ships as part of NumPy as numpy.f2py.

  3. 2001

    Co-creating SciPy

    In 2001, he merges code with Travis Oliphant and Eric Jones into a package they call SciPy. It supplies a standard collection of common numerical operations on top of the Numeric array data structure.

  4. By 2011

    Simula and SymPy Core

    By 2011, he is a research scientist at Simula Research Laboratory in Oslo, Norway, developing scientific software, while on leave from his post as a senior researcher at CENS in Tallinn, Estonia. He is the author of the SymPy Core project and a NumPy developer mainly responsible for the numpy.distutils and numpy.f2py packages.

  5. 2020

    SciPy 1.0 and NumPy papers

    In 2020, he is among the authors of "SciPy 1.0: fundamental algorithms for scientific computing in Python" in Nature Methods. He also co-authors "Array programming with NumPy", published in Nature in September 2020 and listing his affiliation as Quansight in Austin, TX. The paper presents NumPy as the primary array programming library for Python.

  6. 2021–2022

    Sparse tensors in PyTorch

    In April 2021, he proposes a roadmap for completing the CSR layout for PyTorch tensors, which makes matrix multiplication with sparse matrices ten-fold faster. He follows it with a post on masked operations on sparse tensors in March 2022 and one on BSR tensor invariants in June 2022.

  7. As of October 2026

    Staff engineer at OpenTeams

    He is a Staff Software Engineer at OpenTeams, with interests in functional algorithms for complex and real valued functions, validated across libraries such as NumPy, PyTorch and JAX, and in optimizing Triton kernel parameters. The NumPy team lists him among its maintainers, and in January 2026 he publishes a blog post on implementing PyTorch autograd backward methods for tensor operations.

Career

  1. NowOpenTeamsStaff Software Engineer
  2. Simula Research LaboratoryResearch Scientist
  3. CENSSenior Researcher

From the biography.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  2. The United States employs an estimated 48% of people shaping AIWe estimate where the people shaping AI work today, how universities, companies and research institutes split across countries, and how country shares have changed since the 1960s.
  3. 43% of living people shaping AI work outside their country of birthWe estimate that 43% of living people shaping AI work outside their country of birth (520 of 1,198 people with known birth and employment countries).
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionWe 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.

Sources and credits

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

  1. Pearu Peterson · sympy/sympy Wiki · GitHub github.com. Checked 22 October 2026.
  2. ANN: F2PY - Fortran to Python Interface Generator - SciPy-Dev - python.org mail.python.org. Checked 22 October 2026.
  3. F2PY user guide and reference manual — NumPy v2.5 Manual numpy.org. Checked 22 October 2026.
  4. Fortran to Python Interface Generator with an Application to Aerospace Engineering · MDO Lab mdolab.engin.umich.edu. Checked 22 October 2026.
  5. SciPy - Wikipedia en.wikipedia.org. Checked 22 October 2026.
  6. Array programming with NumPy | Nature nature.com. Checked 22 October 2026.
  7. Virtanen, P., Gommers, R., Oliphant, T.E., Peterson, P. et al. - TalTech Research Portal ws.lib.ttu.ee. Checked 22 October 2026.
  8. Pearu’s blogs | Pearu Peterson pearu.github.io. Checked 22 October 2026.
  9. CSR tensor support in PyTorch | Pearu Peterson pearu.github.io. Checked 22 October 2026.
  10. pearu (Pearu Peterson) · GitHub github.com. Checked 22 October 2026.
  11. NumPy - NumPy Teams numpy.org. Checked 22 October 2026.
  12. Implementing PyTorch autograd backward methods for tensor operations | Pearu Peterson pearu.github.io. Checked 22 October 2026.