nvidia-cublas-cu11

What it is and what it does

nvidia-cublas-cu11 is a runtime library package that bundles NVIDIA's CUBLAS (CUDA Basic Linear Algebra Subroutines) native binaries for CUDA 11. It provides no Python API of its own; instead, it serves as a dependency for machine learning and scientific computing frameworks that need GPU-accelerated linear algebra operations. The package distributes precompiled binaries for Linux (x86_64 and ARM64) and Windows (x86_64), allowing downstream packages to locate and load the CUBLAS runtime at install time rather than requiring users to manually install CUDA.

The package is abandoned, with its last release over 1396 days ago and no active repository or maintenance. It carries an unclear proprietary license from NVIDIA. Users typically encounter this as a transitive dependency of PyTorch, TensorFlow, or similar frameworks rather than installing it directly. Because it is no longer maintained and CUDA tooling has evolved, new projects should evaluate whether to use this version or migrate to newer CUDA runtime packages.

Use it for:

• Satisfy GPU linear algebra dependencies for PyTorch or TensorFlow installations targeting CUDA 11 environments.
• Enable GPU-accelerated matrix operations in scientific computing workflows on x86_64 or ARM64 Linux systems.
• Provide CUBLAS runtime binaries for Windows-based deep learning development without manual CUDA toolkit installation.
• Support legacy projects locked to CUDA 11 that require explicit CUBLAS runtime availability.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides NVIDIA CUBLAS native runtime libraries for CUDA 11, enabling GPU-accelerated linear algebra operations in Python applications on x86_64 and ARM64 Linux, Windows platforms.

No for new projects. This package is abandoned (last release 1396 days ago) and carries an unclear proprietary license. Newer CUDA runtime packages and updated frameworks have superseded it. Install only if you are maintaining legacy code explicitly pinned to CUDA 11 and cannot upgrade; otherwise, use current nvidia-cublas packages or let your framework (PyTorch, TensorFlow) manage CUDA runtime dependencies.

Install

nvidia-cublas-cu11 on PyPI

pip install nvidia-cublas-cu11

uv add nvidia-cublas-cu11

poetry add nvidia-cublas-cu11

Installing nvidia-cublas-cu11

Before you install

Medium install friction due to platform-specific wheel distributions (manylinux1/2014 x86_64, aarch64, Windows). Package is abandoned as of 1396 days since last release; no active maintenance or repository commits available.

License in practice

Licensed under NVIDIA Proprietary Software with unclear license treatment. Users should verify compliance with NVIDIA's terms before deploying in production or commercial contexts.

Quickstart

pip install nvidia-cublas-cu11==11.11.3.6

import nvidia.cublas.cu11
# Use via downstream packages (e.g., PyTorch, TensorFlow) that depend on this runtime

Requires CUDA 11 compatible GPU and appropriate NVIDIA drivers; only distributes for Linux (x86_64, aarch64) and Windows (x86_64).

Verify before relying

• Whether this package remains compatible with current NVIDIA driver versions and CUDA 11 ecosystem.
• Specific license restrictions on commercial or redistributed use under NVIDIA Proprietary Software terms.
• Whether downstream packages (PyTorch, TensorFlow) still accept or recommend this version for new projects.

Package facts

Evidence: nvidia_cublas_cu11-11.11.3.6-py3-none-manylinux1_x86_64.whl; nvidia_cublas_cu11-11.11.3.6-py3-none-manylinux2014_aarch64.whl; nvidia_cublas_cu11-11.11.3.6-py3-none-manylinux2014_x86_64.whl; nvidia_cublas_cu11-11.11.3.6-py3-none-win_amd64.whl

Keywords: cuda, nvidia, runtime, machine learning, deep learning

Tags

More Software Development packages

Further reading