NVIDIA ships CUDA Python 1.0 with stable APIs, making Python first-class for CUDA

PyTorch · x · 2026-09-09

Alongside CUDA 13.3, NVIDIA released CUDA Python 1.0, establishing Python as a supported, first-class way to use CUDA with semantic versioning. cuda.core 1.0.0 offers a Pythonic runtime interface (devices, streams, green contexts, checkpointing), cuda.compute 1.0.0 exposes CCCL parallel algorithms with Python lambdas, and cuda.bindings 13.3.0 covers CUDA C APIs 1:1. PyTorch, CuPy, and NVIDIA's own libraries now build on the shared cuda.core foundation, reducing version conflicts.

Original post →

More from Infra

Infra channel →