nvidia-curand-cu11

What it is and what it does

This package distributes NVIDIA's CURAND native runtime libraries as Python wheels, allowing developers to access GPU-accelerated random number generation via CUDA 11. It is a binary distribution package—not a Python API wrapper—that installs compiled NVIDIA libraries for use by downstream CUDA-aware Python packages.

The package targets scientific computing, machine learning, and deep learning workflows where reproducible, high-performance random sampling on GPUs is needed. It supports Python 3.5 through 3.11 across Linux (x86_64 and ARM64) and Windows platforms, but has been abandoned since 2022-10-03 with no ongoing maintenance or updates.

Use it for:

• Install as a dependency for CUDA libraries that require CURAND runtime on systems with CUDA 11.
• Enable GPU-accelerated random number generation in machine learning training pipelines that call CURAND directly.
• Provide CURAND binaries in containerized environments (Docker) where CUDA 11 is the base runtime.
• Support legacy deep learning projects pinned to CUDA 11 that cannot upgrade to newer CUDA versions.

Worth the install?

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

Provides NVIDIA CURAND native runtime libraries for CUDA 11, enabling GPU-accelerated random number generation in Python applications on x86_64 and ARM64 Linux, Windows, and legacy x86 platforms.

No—install only if you are locked into CUDA 11 and have no path to upgrade. The package is abandoned (last release 2022-10-03), receives no maintenance or security updates, and its proprietary license terms are unclear. For new projects, use current CUDA versions with actively maintained runtime packages. For existing CUDA 11 deployments, verify compatibility before committing.

Install

nvidia-curand-cu11 on PyPI

pip install nvidia-curand-cu11

uv add nvidia-curand-cu11

poetry add nvidia-curand-cu11

Installing nvidia-curand-cu11

Before you install

Medium install friction due to platform-specific wheel distribution (x86_64, ARM64, Windows variants). Package is abandoned—last release was 2022-10-03, with no maintenance, security updates, or compatibility fixes since then.

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-curand-cu11

import curand
# Use curand functions for GPU random number generation

Requires CUDA 11 toolkit installed on the system; only works on x86_64/ARM64 Linux or Windows; Python >=3 required.

Verify before relying

• Whether this package remains compatible with current CUDA 11 runtime versions given its abandoned status.
• Clarification on NVIDIA Proprietary Software license terms and redistribution/commercial use restrictions.
• Whether newer CUDA versions (12.x) are supported or if users must pin to CUDA 11.

Package facts

Evidence: nvidia_curand_cu11-10.3.0.86-py3-none-manylinux1_x86_64.whl; nvidia_curand_cu11-10.3.0.86-py3-none-manylinux2014_aarch64.whl; nvidia_curand_cu11-10.3.0.86-py3-none-manylinux2014_x86_64.whl; nvidia_curand_cu11-10.3.0.86-py3-none-win_amd64.whl

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

Tags

More Software Development packages

Further reading