<ul data-eligibleForWebStory="false"><li>NVIDIA has developed Python SDKs to leverage modern GPUs for data science and ML with nearly native performance.</li><li>Bryce Adelstein Lelbach discusses programming GPUs in pure Python on Talk Python episode #509.</li><li>Sponsors of the episode include Posit, Agntcy, and Talk Python Courses.</li><li>Links mentioned in the show: Bryce Adelstein Lelbach on Twitter, NVIDIA CUDA Python API, Numba, Applied Data Science Podcast, NVIDIA Accelerated Computing Hub, CUDA Cooperative Groups, Numba CUDA User Guide, CuPy, RAPIDS User Guide, and more.</li><li>Resources such as Google Colab, Compiler Explorer, and NVIDIA's First Desktop AI PC were shared.</li><li>Viewers can find additional resources, deep-dive content, and episode transcripts on the Talk Python website.</li><li>Ways to stay in touch include subscribing to Talk Python on YouTube, following Talk Python and Michael Kennedy on Bluesky and Mastodon.</li></ul>

Talk Python to Me: #509: GPU Programming in Pure Python

Discover more