nvidia-nccl-cu12
NVIDIA Collective Communication Library (NCCL) Runtime
Decision gist · record as of 2026-08-14
Yes, if you are running distributed GPU workloads on CUDA 12 hardware. This is a foundational runtime library for multi-GPU communication; most modern distributed deep learning frameworks depend on it. Install only if your system has compatible NVIDIA GPUs and CUDA 12 already installed. No known vulnerabilities and active maintenance are positive signals.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires CUDA 12-compatible NVIDIA GPU hardware and compatible Linux or Windows system; wheels available only for manylinux_2_18 x86_64 and aarch64
- Medium install friction due to platform-specific wheels (manylinux_2_18 x86_64 and aarch64 only); active maintenance with a release 3 days ago suggests ongoing support.
License · maintenance · safety
(unclear)
last release 2026-08-11 (3 days)
0 known vulnerabilities (OSV.dev, 2026-08-14) · 52,431,980 downloads/mo, #560 on PyPI
Alternatives
Verify before relying
pip install nvidia-nccl-cu12==2.31.2
# Typically imported and used indirectly by distributed frameworks
import nvidia.nccl- Whether CUDA 12 must be installed separately on the system for this package to function
- Whether InfiniBand or TCP/IP fallback is automatic or requires explicit configuration
- Minimum GPU hardware requirements beyond CUDA 12 support
What it is and what it does
nvidia-nccl-cu12 is NVIDIA's Collective Communication Library runtime for CUDA 12, packaged for Python environments. It provides optimized implementations of standard multi-GPU communication patterns—all-reduce, all-gather, reduce, broadcast, and reduce-scatter—designed to achieve high bandwidth across PCIe, NVLink, NVswitch interconnects, and network fabrics using InfiniBand Verbs or TCP/IP sockets.
This is a runtime library typically used as a dependency by distributed deep learning frameworks rather than called directly. The package is actively maintained and widely used in machine learning workloads requiring efficient multi-GPU communication.
Use it for
- Distributed training of neural networks across multiple GPUs on a single machine or cluster
- Synchronizing gradients and model parameters in data-parallel deep learning frameworks
- Implementing custom collective operations in GPU-accelerated scientific computing applications
- Enabling efficient all-reduce and broadcast operations in custom CUDA kernels
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are running distributed GPU workloads on CUDA 12 hardware.
This is a foundational runtime library for multi-GPU communication; most modern distributed deep learning frameworks depend on it. Install only if your system has compatible NVIDIA GPUs and CUDA 12 already installed. No known vulnerabilities and active maintenance are positive signals.
Install
nvidia-nccl-cu12 on PyPI
Before you install
Medium install friction due to platform-specific wheels (manylinux_2_18 x86_64 and aarch64 only); active maintenance with a release 3 days ago suggests ongoing support.
Requires CUDA 12-compatible NVIDIA GPU hardware and compatible Linux or Windows system; wheels available only for manylinux_2_18 x86_64 and aarch64
Quickstart
pip install nvidia-nccl-cu12==2.31.2
# Typically imported and used indirectly by distributed frameworks
import nvidia.ncclVerify before relying
- Whether CUDA 12 must be installed separately on the system for this package to function
- Whether InfiniBand or TCP/IP fallback is automatic or requires explicit configuration
- Minimum GPU hardware requirements beyond CUDA 12 support
Package facts
| License | Not declared unclear |
| Python support | Supports the current Python release >=3 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | None |
| Maintenance | Actively maintained 3 days since the last release |
| First released | |
| Downloads | 52,431,980 / month, #560 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 4 - BetaIntended Audience :: DevelopersIntended Audience :: EducationIntended Audience :: Science/ResearchNatural Language :: EnglishOperating System :: Microsoft :: WindowsOperating System :: POSIX :: LinuxProgramming Language :: Python :: 3Programming Language :: Python :: 3 :: OnlyProgramming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.5Programming Language :: Python :: 3.6Programming Language :: Python :: 3.7Programming Language :: Python :: 3.8Programming Language :: Python :: 3.9Topic :: Scientific/EngineeringTopic :: Scientific/Engineering :: Artificial IntelligenceTopic :: Scientific/Engineering :: MathematicsTopic :: Software DevelopmentTopic :: Software Development :: Libraries |
Evidence: nvidia_nccl_cu12-2.31.2-py3-none-manylinux_2_18_aarch64.whl; nvidia_nccl_cu12-2.31.2-py3-none-manylinux_2_18_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “gpu all-reduce broadcast”
- nvidia-nccl-cu12Provides NVIDIA's NCCL runtime library for GPU collective…
- nvidia-nccl-cu13Provides NVIDIA's Collective Communication Library (NCCL) runtime for…
- nvidia-nccl-cu11Provides NVIDIA's collective communication library (NCCL) runtime for…
Give your agent the search over MCP, or paste the wish link into any chat.
More Software Development packages
Provides backported and experimental type hints for Python 3.9+, allowing use of newer typing features on older Python versions and enabling early experimentation with type system PEPs before they enter the standard library.
NumPy provides an N-dimensional array object and a comprehensive suite of mathematical, linear algebra, Fourier transform, and random number functions for scientific computing in Python.
FastAPI is a Python web framework for building REST APIs using type hints, with automatic request validation, serialization, and interactive API documentation.
Provides a way to document function parameters, class attributes, return types, and variables inline using Python's `Annotated` type hint syntax instead of traditional docstrings.
Typer builds command-line applications from Python functions using type hints, automatically generating help text, argument parsing, and shell completion.
Install it if you are building CLIs in Python.
Distlib provides low-level packaging utilities for building, distributing, and managing Python software—including metadata handling, version specifiers, wheel support, script installation, and dependency resolution.
See also nvidia-nccl-cu11 · nvidia-nccl-cu13 · nccl4py · nvidia-nvshmem-cu12 · distributed-ucxx-cu12 · nvidia-nvshmem-cu13 · nvidia-cuda-cccl · nvidia-cuda-cccl-cu12 · nvidia-cuda-runtime-cu12 · lcm