Skip to main content
Ctrl+K
NVIDIA RAPIDS Documentation - Home NVIDIA RAPIDS Documentation - Home

NVIDIA RAPIDS Documentation

  • GitHub
NVIDIA RAPIDS Documentation - Home NVIDIA RAPIDS Documentation - Home

NVIDIA RAPIDS Documentation

  • GitHub

Table of Contents

  • RAPIDS Installation Guide
  • RAPIDS Platform Support
  • RAPIDS Featured User Guides
  • RAPIDS API Documentation
  • RAPIDS Visualization Guide
  • RAPIDS Maintainers Docs
    • Datasets
    • Forward Mergers
    • Releases
      • Release Process
      • Release Schedule
      • Hotfix Process
    • Resources
      • Contributor Covenant Code of Conduct
      • Changelog Format
      • Auto-Merger
      • GitHub Actions
      • Reproducing CI Locally
      • Telemetry
      • Label Checker
      • Merge Barriers Check
      • Recently Updated Check
      • Versions and Tags
      • Burn down guide
  • Contributing
    • Issues
    • Code Contributions
    • Pull Requests
  • RAPIDS Notices
    • RAPIDS Developer Notices
    • RAPIDS General Notices
    • RAPIDS Support Notices
  • RSN 52 - Dropping of pynvjitlink in RAPIDS Release v25.10

RSN 52 - Dropping of pynvjitlink in RAPIDS Release v25.10#

Author

RAPIDS Ops

Status

Completed

Topic

Library Deprecation

RAPIDS Version

v25.10

Created

August 18, 2025

Updated

N/A

Overview#

Starting with RAPIDS v25.10, RAPIDS libraries no longer depend on pynvjitlink.

The final release of pynvjitlink was v0.7.0. Future releases are not planned, and rapidsai/pynvjitlink will be archived.

Impact#

If you are only using pynvjitlink indirectly through RAPIDS libraries, no code changes are required.

The pynvjitlink library has been discontinued in favor of functionality in cuda.core.experimental.Linker (docs). Users of numba-cuda>=0.16 can eliminate uses of from pynvjitlink import patch_numba_linker; patch_numba_linker(). The numba.cuda target will now automatically detect and enable nvjitlink when needed with no explicit configuration.

All other public APIs of pynvjitlink are replaced by cuda.core and cuda.bindings functionality.

Documentation:

  • cuda.core: https://nvidia.github.io/cuda-python/cuda-core/latest/

  • cuda.bindings: https://nvidia.github.io/cuda-python/cuda-bindings/latest/

  • numba-cuda: https://nvidia.github.io/numba-cuda/

On this page
  • Overview
  • Impact
NVIDIA NVIDIA
Privacy Policy | Your Privacy Choices | Terms of Service | Accessibility | Corporate Policies | Product Security | Contact

Copyright © 2026, NVIDIA.