#544: Wheel Next + Packaging PEPs
When you pip install a package with compiled code, the wheel you get is built for CPU features from 2009. Want newer optimizations like AVX2? Your installer has no way to ask for them. GPU support? You're on your own configuring special index URLs. The result is fat binaries, nearly gigabyte-sized wheels, and install pages that read like puzzle books. A coalition from NVIDIA, Astral, and QuanSight has been working on Wheel Next: A set of PEPs that let packages declare what hardware they need and let installers like uv pick the right build automatically. Just uv pip install torch and it works. I sit down with Jonathan Dekhtiar from NVIDIA, Ralf Gommers from Quansight and the NumPy and SciPy teams, and Charlie Marsh, founder of Astral and creator of uv, to dig into all of it.
Episode sponsors
Sentry Error Monitoring, Code talkpython26
Temporal
Talk Python Courses
Links from the show
Guests
Charlie Marsh: github.com
Ralf Gommers: github.com
Jonathan Dekhtiar: github.com
CPU dispatcher: numpy.org
build options: numpy.org
Red Hat RHEL: www.redhat.com
Red Hat RHEL AI: www.redhat.com
RedHats presentation: wheelnext.dev
CUDA release: developer.nvidia.com
requires a PEP: discuss.python.org
WheelNext: wheelnext.dev
Github repo: github.com
PEP 817: peps.python.org
PEP 825: discuss.python.org
uv: docs.astral.sh
A variant-enabled build of uv: astral.sh
pyx: astral.sh
pypackaging-native: pypackaging-native.github.io
PEP 784: peps.python.org
Watch this episode on YouTube: youtube.com
Episode #544 deep-dive: talkpython.fm/544
Episode transcripts: talkpython.fm
Theme Song: Developer Rap
🥁 Served in a Flask 🎸: talkpython.fm/flasksong
---== Don't be a stranger ==---
YouTube: youtube.com/@talkpython
Bluesky: @talkpython.fm
Mastodon: @talkpython@fosstodon.org
X.com: @talkpython
Michael on Bluesky: @mkennedy.codes
Michael on Mastodon: @mkennedy@fosstodon.org
Michael on X.com: @mkennedy
Open-source is for the people, by the people
Travis Oliphant, creator of NumPy and SciPy, joins Ryan to explore the development of Python as a data science tool, the evolution of these foundational libraries, and the importance of community and collaboration in open-source projects, including Travis’ current work to support sustainable open-source through the OpenTeams Incubator.
Episode notes:
NumPy and SciPy are the fundamental packages and algorithms for scientific computing with Python. NumPy 2.3.0 and SciPy 1.16.0 are out now.
The OpenTeams Incubator helps start, grow, and sustain open-source software communities.
Quansight is a data, science, and engineering firm rooted in the work of the Python Data, Science, and AI/ML open-source communities.
Connect with Travis on LinkedIn or email him at travis@OTincubator.com
Today we’re shouting out user RobinFrcd for answering pytest-asyncio has a closed event loop, but only when running all tests and winning a Populist badge.
See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.
Python documentary companion pod (Interview)
Our friends at Cult.Repo launched their epic Python documentary on August 28th, 2025! To celebrate, we sat down with Travis Oliphant –creator of NumPy, SciPy, and more– to get his perspective on how Python took over the software world.
Stick around for the twist ending! We set aside Python and dissect Travis’ big idea to make open source projects financially sustainable through direct investment.
Join the discussion
Changelog++ members save 4 minutes on this episode because they made the ads disappear. Join today!
Sponsors:
Depot – 10x faster builds? Yes please. Build faster. Waste less time. Accelerate Docker image builds, and GitHub Actions workflows. Easily integrate with your existing CI provider and dev workflows to save hours of build time.
Auth0 – The identity infrastructure for the age of AI. Built by developers, for developers—Auth0 helps you secure users, agents, and third-party access across modern AI workflows. Token vaulting, fine-grained authorization, and standards-based auth, all in one platform.
Start building at Auth0.com/ai
Fly.io – The home of Changelog.com — Deploy your apps close to your users — global Anycast load-balancing, zero-configuration private networking, hardware isolation, and instant WireGuard VPN connections. Push-button deployments that scale to thousands of instances. Check out the speedrun to get started in minutes.
Featuring:
Travis Oliphant – GitHub, LinkedIn, Mastodon, X
Adam Stacoviak – Website, GitHub, LinkedIn, Mastodon, X
Jerod Santo – Website, GitHub, LinkedIn, Mastodon, X
Show Notes:
Python: The Documentary
NumPy
SciPy.org
Mojo 🔥: Powerful CPU+GPU Programming
FairOSS
tea.xyz
Cult.Repo
Something missing or broken? PRs welcome!
Travis Oliphant: SciPy, NumPy, and Fostering Scientific Python
<p>What went into developing the open-source Python tools data scientists use every day? This week on the show, we talk with Travis Oliphant about his work on SciPy, NumPy, Numba, and many other contributions to the Python scientific community.</p>
<p>Travis discusses his initial involvement in the open-source community and how he discovered Python while working in biomedical imaging. He was trying to find ways to manage large sets of numerical data, which led to his initial contributions and collaborations in building scientific libraries. </p>
<p>His appearance on the show coincides with the release of the Python documentary, in which he’s featured. We discuss the myriad organizations Travis founded, including Quansight, OpenTeams, and Anaconda. We dig into his underlying mission to continue fostering the growth of the open-source scientific computing community.</p>
<p>This episode is sponsored by InfluxData.</p>
<div class="alert alert-primary" role="alert">
<p><strong>Course Spotlight:</strong> <a href="https://realpython.com/courses/numpy-techniques-practical-examples/">NumPy Techniques and Practical Examples</a></p>
<p>In this video course, you’ll learn how to use NumPy by exploring several interesting examples. You’ll read data from a file into an array and analyze structured arrays to perform a reconciliation. You’ll also learn how to quickly chart an analysis and turn a custom function into a vectorized function.</p>
</div>
<p>Topics:</p>
<ul>
<li>00:00:00 – Introduction</li>
<li>00:02:41 – Python documentary</li>
<li>00:07:44 – Getting involved in open source</li>
<li>00:12:04 – Numeric Python</li>
<li>00:15:36 – SciPy and the SciPy community</li>
<li>00:17:35 – Starting to think about entrepreneurship </li>
<li>00:18:16 – NumPy evolving from the work of Numeric</li>
<li>00:22:01 – Sponsor: InfluxData</li>
<li>00:22:53 – Python as controlling code for lower-level libraries</li>
<li>00:23:37 – Numba open-source JIT compiler</li>
<li>00:30:09 – Starting to build in Python before learning it all</li>
<li>00:34:45 – Python as the language AI generates</li>
<li>00:36:31 – Guilds and sharing knowledge</li>
<li>00:40:15 – More NumPy backstory</li>
<li>00:46:36 – Contributing to Python</li>
<li>00:48:24 – Video Course Spotlight</li>
<li>00:49:41 – The investment of companies in Python</li>
<li>00:51:22 – Quansight and businesses in open source</li>
<li>00:53:09 – Open Teams and Quansight details</li>
<li>00:57:14 – NumFOCUS and Anaconda</li>
<li>00:58:51 – FairOSS</li>
<li>01:02:36 – Documenting these efforts</li>
<li>01:05:37 – What are you excited about in the world of Python?</li>
<li>01:07:12 – What do you want to learn next?</li>
<li>01:08:10 – How can people follow your work online?</li>
<li>01:10:03 – Thanks and goodbye</li>
</ul>
<p>Show Links:</p>
<ul>
<li><a href="https://www.youtube.com/watch?v=pqBqdNIPrbo">Python: The Documentary - OFFICIAL TRAILER - Coming August 28 - YouTube</a></li>
<li><a href="http://hugunin.net/papers/hugunin95numpy.html">The Python Matrix Object: Extending Python for Numerical Computation</a></li>
<li><a href="https://j1m.dev/">Jim Fulton</a></li>
<li><a href="http://hugunin.net/">Jim Hugunin - Home</a></li>
<li><a href="https://scipy.github.io/old-wiki/pages/History_of_SciPy">History of SciPy - SciPy wiki dump</a></li>
<li><a href="https://scipy.org/">SciPy</a></li>
<li><a href="https://numpy.org/">NumPy</a></li>
<li><a href="https://numba.pydata.org/">Numba: A High Performance Python Compiler</a></li>
<li><a href="https://lpython.org/">LPython - High performance typed Python compiler</a></li>
<li><a href="https://openteams.com/">OpenTeams: Open SaaS AI Solutions</a></li>
<li><a href="https://quansight.com/">Quansight Consulting</a></li>
<li><a href="https://otincubator.com/">OpenTeams Incubator</a></li>
<li><a href="https://numfocus.org/">NumFOCUS: A Nonprofit Supporting Open Code for Better Science</a></li>
<li><a href="https://anaconda.com/">Anaconda</a></li>
<li><a href="https://faiross.org/">FairOSS</a></li>
<li><a href="https://github.com/faster-cpython/">faster-cpython</a></li>
<li><a href="https://x.com/teoliphant">Travis Oliphant (@teoliphant) / X</a></li>
<li><a href="https://www.linkedin.com/in/teoliphant/">Travis Oliphant - LinkedIn</a></li>
</ul>
<p>Level up your Python skills with our expert-led courses:</p>
<ul>
<li><a href="https://realpython.com/courses/data-cleaning-with-pandas-and-numpy/">Data Cleaning With pandas and NumPy</a></li>
<li><a href="https://realpython.com/courses/stacks-queues-ideal-data-structure/">Stacks and Queues: Selecting the Ideal Data Structure</a></li>
<li><a href="https://realpython.com/courses/numpy-techniques-practical-examples/">NumPy Techniques and Practical Examples</a></li>
</ul> <p><a rel="payment" href="https://realpython.com/join">Support the podcast & join our community of Pythonistas</a></p>
765: NumPy, SciPy and the Economics of Open-Source, with Dr. Travis Oliphant
Explore the origins of NumPy and SciPy with their creator, Dr. Travis Oliphant. Discover the journey from personal need to global impact, the challenges overcome, and the future of these essential Python libraries in scientific computing and data science.
This episode is brought to you by the DataConnect Conference, by Data Universe, the out-of-this-world data conference, and by CloudWolf, the Cloud Skills platform. Interested in sponsoring a SuperDataScience Podcast episode? Visit passionfroot.me/superdatascience for sponsorship information.
In this episode you will learn:
• Travis's journey to creating NumPy and SciPy [08:05]
• How Anaconda got started [42:24]
• How Numba, a high-performance Python compiler, was brought to market [54:48]
• Python's influence on the thought processes of scientists and engineers [1:04:21]
• The commercial projects that support Travis’s vast open-source efforts and communities [1:10:22]
• How to get involved in Travis's commercial projects and communities [1:22:34]
• The future of scientific computing and Python libraries [1:29:50]
Additional materials: www.superdatascience.com/765
Exploring pandas 2.0 & Targets for Apache Arrow
<p>What are the new ways to describe your data in pandas 2.0? Will the addition of Apache Arrow to the data back end foster the growth of data interoperability? This week on the show, we talk with pandas core developer Marc Garcia about the release of pandas 2.0.</p>
<p>Marc shares his background and work on pandas. We discuss the history of data representation in pandas and the need to move beyond NumPy. We also talk about how Apache Arrow only solves some of the issues.</p>
<p>We dig into the potential of an Apache Arrow back end and how it could offer interoperability between data platforms. We also cover the moderate adoption and backward-compatibility concerns. Marc also shares his thoughts on making pandas more extensible.</p>
<div class="alert alert-primary" role="alert">
<p><strong>Course Spotlight:</strong> <a href="https://realpython.com/courses/pandas-dataframe-working-with-data/">The pandas DataFrame: Working With Data Efficiently</a> </p>
<p>In this course, you’ll get started with pandas DataFrames, which are powerful and widely used two-dimensional data structures. You’ll learn how to perform basic operations with data, handle missing values, work with time-series data, and visualize data from a pandas DataFrame.</p>
</div>
<p>Topics:</p>
<ul>
<li>00:00:00 – Introduction</li>
<li>00:02:07 – Getting involved with the pandas project</li>
<li>00:03:48 – Continued growth of the platform</li>
<li>00:06:49 – Parallel branch development</li>
<li>00:09:19 – The introduction of Apache Arrow</li>
<li>00:18:53 – Working with NumPy data in pandas</li>
<li>00:30:18 – Arrow data types and strings</li>
<li>00:41:23 – Video Course Spotlight</li>
<li>00:42:37 – Interoperability of Arrow data back end</li>
<li>00:50:36 – Could pandas be more extensible?</li>
<li>01:00:49 – Python DataFrame Summit 2023</li>
<li>01:08:12 – What are you excited about in the world of Python?</li>
<li>01:11:13 – What do you want to learn next?</li>
<li>01:12:12 – How can people follow your work online?</li>
<li>01:13:46 – Thanks and Goodbye</li>
</ul>
<p>Show Links:</p>
<ul>
<li><a href="https://datapythonista.me/">Marc Garcia - datapythonista - data engineer, data scientist and pandas core developer</a></li>
<li><a href="https://datapythonista.me/blog/pandas-20-and-the-arrow-revolution-part-i">pandas 2.0 and the Arrow revolution (part I)</a></li>
<li><a href="https://tubedu.org/w/0054e518-804e-4d91-9388-11d7008fd909">The pandas of the future - Marc Garcia - SciPyLA 2019 - TubEdu</a></li>
<li><a href="https://slate.com/technology/2019/10/round-floor-software-errors-stock-market-battlefield.html">The deadly consequences of rounding errors - Slate</a></li>
<li><a href="https://pandas.pydata.org/community/blog/">Community Blog - pandas - Python Data Analysis Library</a></li>
<li><a href="https://arrow.apache.org/">Apache Arrow - Apache Arrow</a></li>
<li><a href="https://wesmckinney.com/blog/apache-arrow-pandas-internals/">Apache Arrow and the “10 Things I Hate About pandas” - Wes McKinney</a></li>
<li><a href="https://pandas.pydata.org/pdeps/0009-io-extensions.html">I/O Extensions in pandas - PDEP-9</a></li>
<li><a href="http://tomaugspurger.net/posts/pandas-extension-arrays/">Extension Arrays for Pandas - Tom’s Blog</a></li>
<li><a href="https://df-summit-23.pydata.org/">Python Dataframe Summit 2023</a></li>
<li><a href="https://www.rust-lang.org/">Rust Programming Language</a></li>
<li><a href="https://en.wikipedia.org/wiki/Freediving">Freediving - Wikipedia</a></li>
<li><a href="https://www.linkedin.com/in/datapythonista/">Marc Garcia - LinkedIn</a></li>
<li><a href="https://twitter.com/datapythonista">Marc Garcia (@datapythonista) - X</a></li>
</ul>
<p>Level up your Python skills with our expert-led courses:</p>
<ul>
<li><a href="https://realpython.com/courses/reading-writing-files-pandas/">Reading and Writing Files With pandas</a></li>
<li><a href="https://realpython.com/courses/explore-dataset-with-pandas/">Explore Your Dataset With pandas</a></li>
<li><a href="https://realpython.com/courses/pandas-dataframe-working-with-data/">The pandas DataFrame: Working With Data Efficiently</a></li>
</ul> <p><a rel="payment" href="https://realpython.com/join">Support the podcast & join our community of Pythonistas</a></p>
#224 – Travis Oliphant: NumPy, SciPy, Anaconda, Python & Scientific Programming
Travis Oliphant is a data scientist, entrepreneur, and creator of NumPy, SciPy, and Anaconda. Please support this podcast by checking out our sponsors:
– Novo: https://banknovo.com/lex
– Allform: https://allform.com/lex to get 20% off
– Onnit: https://lexfridman.com/onnit to get up to 10% off
– Athletic Greens: https://athleticgreens.com/lex and use code LEX to get 1 month of fish oil
– Blinkist: https://blinkist.com/lex and use code LEX to get 25% off premium
EPISODE LINKS:
Travis’s Twitter: https://twitter.com/teoliphant
Travis’s Wiki Page: https://en.wikipedia.org/wiki/Travis_Oliphant
NumPy: https://numpy.org/
SciPy: https://scipy.org/about.html
Anaconda: https://www.anaconda.com/products/individual
Quansight: https://www.quansight.com
PODCAST INFO:
Podcast website: https://lexfridman.com/podcast
Apple Podcasts: https://apple.co/2lwqZIr
Spotify: https://spoti.fi/2nEwCF8
RSS: https://lexfridman.com/feed/podcast/
YouTube Full Episodes: https://youtube.com/lexfridman
YouTube Clips: https://youtube.com/lexclips
SUPPORT & CONNECT:
– Check out the sponsors above, it’s the best way to support this podcast
– Support on Patreon: https://www.patreon.com/lexfridman
– Twitter: https://twitter.com/lexfridman
– Instagram: https://www.instagram.com/lexfridman
– LinkedIn: https://www.linkedin.com/in/lexfridman
– Facebook: https://www.facebook.com/lexfridman
– Medium: https://medium.com/@lexfridman
OUTLINE:
Here’s the timestamps for the episode. On some podcast players you should be able to click the timestamp to jump to that time.
(00:00) – Introduction
(07:06) – Early programming
(28:47) – SciPy
(45:41) – Open source
(57:23) – NumPy
(1:34:39) – Guido van Rossum
(1:46:57) – Efficiency
(1:55:49) – Objects
(2:02:47) – Numba
(2:11:53) – Anaconda
(2:16:20) – Conda
(2:31:56) – Quansight Labs
(2:35:32) – OpenTeams
(2:43:05) – GitHub
(2:48:35) – Marketing
(2:53:13) – Great programming
(3:04:03) – Hiring
(3:08:01) – Advice for young people
#217: Notebooks vs data science-enabled scripts
See the full show notes for this episode on the website at talkpython.fm/217