News

NVIDIA Partners with Third-Party Platforms to Distribute CUDA Software

• By John K. Waters
• September 10, 2025

NVIDIA has announced partnerships with several operating system providers and package managers to redistribute its CUDA parallel computing platform, aiming to simplify software deployment for developers working with GPU-accelerated applications.

The graphics chip maker said Monday it is working with Canonical, CIQ, SUSE, and developer environment manager Flox to allow direct distribution of CUDA through their respective platforms. The move addresses longstanding compatibility challenges that developers face when building applications that require GPU processing capabilities.

CUDA, or Compute Unified Device Architecture, is NVIDIA's parallel computing platform that enables developers to use graphics processing units for general-purpose computing tasks. The technology is widely used in artificial intelligence, machine learning, and scientific computing applications.

The new distribution model allows partner platforms to embed CUDA directly into their package management systems, potentially reducing installation complexity for applications like PyTorch and the computer vision library OpenCV that rely on GPU acceleration.

"Building and deploying applications can be challenging for developers, requiring them to navigate the complex relationship between hardware and software capabilities," NVIDIA stated in its announcement.

The company said the initiative supplements existing distribution methods, which include direct downloads, container images, and Python package installations through pip or conda.

Under the agreements, third-party distributors will maintain naming consistency with NVIDIA's conventions and provide timely updates following official releases. NVIDIA emphasized that CUDA software will remain freely available, even when bundled with commercial products, though distributors may charge for their broader software packages or services.

The partnerships represent a shift in NVIDIA's distribution strategy as demand for GPU computing continues to grow across industries. The company said additional distribution partners will be announced in coming months.

Support for CUDA installations through third-party channels will be available through both the distributors and NVIDIA's existing developer forums and documentation.

The announcement comes as enterprises increasingly integrate GPU-accelerated computing into their workflows, particularly for AI and data analytics applications that require parallel processing capabilities.