Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

Reproducible CUDA Accelerated Workflows for Scientists with Pixi

Authors
Affiliations
University of Wisconsin-Madison
prefix.dev GmbH
NVIDIA

Abstract

Scientific researchers need reproducible software environments for complex applications that can run across heterogeneous computing platforms. Modern open source tools, like Pixi, provide automatic reproducibility solutions for all dependencies while providing a high level interface well suited for researchers.

This tutorial will provide a practical introduction to using Pixi to easily create scientific and AI/ML environments that benefit from hardware acceleration, across multiple machines and platforms. The focus will be on CUDA applications, such as machine learning frameworks and use of CUDA Tile, as well as using pixi-build to construct bespoke CUDA enabled conda packages.

Keywords:reproduciblehardware accelerationmachine learningpixicudapythonscipy

Taught at SciPy 2026 as a tutorial on Monday July 13th, 2026

DOI

SciPy Logistical Information

Rough Outline

00:00 – 00:05 (5 min):

00:05 – 00:15 (10 min):

00:15 – 00:30 (15 min):

00:30 – 01:00 (30 min):

01:00 – 01:40 (40 min):

01:40 – 01:55 (15 min):

01:55 – 02:35 (40 min):

02:35 – 02:45 (10 min):

02:45 – 03:15 (30 min):

03:15 – 03:50 (35 min):

03:50 – 04:00 (10 min):

This tutorial was supported by the University of Wisconsin–Madison Data Science Institute, prefix.dev GmbH, and NVIDIA.

Acknowledgments

Special thanks to Tim de Jager and Wolf Vollprecht of prefix.dev GmBH and Daniel Ching of NVIDIA for their review, feedback, and contributions to this tutorial. Special thanks to Nick Hodgskin and Eniola Awowale for volunteering to serve as teaching assistants for this tutorial.

References
  1. Feickert, M., Arts, R., & Riehl, K. (2026). Reproducible CUDA Accelerated Workflows for Scientists with Pixi. Zenodo. 10.5281/ZENODO.21829174