Python’s data stack is getting a serious GPU turbo boost. In this episode, Ben Zaitlen from NVIDIA joins us to unpack RAPIDS, the open source toolkit that lets pandas, scikit-learn, Spark, Polars, and even NetworkX execute on GPUs. We trace the project’s origin and why NVIDIA built it in the open, then dig into the pieces that matter in practice: cuDF for DataFrames, cuML for ML, cuGraph for graphs, cuXfilter for dashboards, and friends like cuSpatial and cuSignal. We talk real speedups, how the pandas accelerator works without a rewrite, and what becomes possible when jobs that used to take hours finish in minutes. You’ll hear strategies for datasets bigger than GPU memory, scaling out with Dask or Ray, Spark acceleration, and the growing role of vector search with cuVS for AI workloads. If you know the CPU tools, this is your on-ramp to the same APIs at GPU speed.

Posit

Talk Python Courses

Links from the show

RAPIDS: github.com/rapidsai

Example notebooks showing drop-in accelerators: github.com

Benjamin Zaitlen - LinkedIn: linkedin.com

RAPIDS Deployment Guide (Stable): docs.rapids.ai

RAPIDS cuDF API Docs (Stable): docs.rapids.ai

Asianometry YouTube Video: youtube.com

cuDF pandas Accelerator (Stable): docs.rapids.ai

Watch this episode on YouTube: youtube.com

Episode #516 deep-dive: talkpython.fm/516

Episode transcripts: talkpython.fm

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @[email protected]

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @[email protected]

Michael on X.com: @mkennedy