NCCL (pronounced "Nickel") is a stand-alone library of standard collective communication routines for GPUs, implementing all-reduce, all-gather, reduce, broadcast, and reduce-scatter. It has been optimized to achieve high bandwidth on any platform using PCIe, NVLink, NVswitch, as well as networking using InfiniBand Verbs or TCP/IP sockets.
Raw data
{
"_id": null,
"home_page": "https://developer.nvidia.com/cuda-zone",
"name": "nvidia-nccl-cu12",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3",
"maintainer_email": null,
"keywords": "cuda, nvidia, runtime, machine learning, deep learning",
"author": "Nvidia CUDA Installer Team",
"author_email": "compute_installer@nvidia.com",
"download_url": null,
"platform": null,
"description": "NCCL (pronounced \"Nickel\") is a stand-alone library of standard collective communication routines for GPUs, implementing all-reduce, all-gather, reduce, broadcast, and reduce-scatter. It has been optimized to achieve high bandwidth on any platform using PCIe, NVLink, NVswitch, as well as networking using InfiniBand Verbs or TCP/IP sockets.\n",
"bugtrack_url": null,
"license": null,
"summary": "NVIDIA Collective Communication Library (NCCL) Runtime",
"version": "2.28.3",
"project_urls": {
"Homepage": "https://developer.nvidia.com/cuda-zone"
},
"split_keywords": [
"cuda",
" nvidia",
" runtime",
" machine learning",
" deep learning"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "9976a87f70dc4544f76abf528c335a5c107130cca5c345725b1d2ad03bf2c2a2",
"md5": "8ad8951a59913762e92fb04ea90b2488",
"sha256": "85144f2197e81148e18f3ffd28a30d78b5046844877630d2710a1b22669a6e46"
},
"downloads": -1,
"filename": "nvidia_nccl_cu12-2.28.3-py3-none-manylinux_2_18_aarch64.whl",
"has_sig": false,
"md5_digest": "8ad8951a59913762e92fb04ea90b2488",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3",
"size": 295887310,
"upload_time": "2025-09-06T00:31:43",
"upload_time_iso_8601": "2025-09-06T00:31:43.228082Z",
"url": "https://files.pythonhosted.org/packages/99/76/a87f70dc4544f76abf528c335a5c107130cca5c345725b1d2ad03bf2c2a2/nvidia_nccl_cu12-2.28.3-py3-none-manylinux_2_18_aarch64.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "5d2411df42593d1a6d10b3ffef049cec064832f108e77bc5cac12726e4ec1cb2",
"md5": "437faab78739ed3f4cd6e846a240cb46",
"sha256": "79cf0412094e4a552889e5cb7757d92c010ead557ec722c5eebe6a94b1d8681c"
},
"downloads": -1,
"filename": "nvidia_nccl_cu12-2.28.3-py3-none-manylinux_2_18_x86_64.whl",
"has_sig": false,
"md5_digest": "437faab78739ed3f4cd6e846a240cb46",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3",
"size": 295901337,
"upload_time": "2025-09-06T00:32:01",
"upload_time_iso_8601": "2025-09-06T00:32:01.348901Z",
"url": "https://files.pythonhosted.org/packages/5d/24/11df42593d1a6d10b3ffef049cec064832f108e77bc5cac12726e4ec1cb2/nvidia_nccl_cu12-2.28.3-py3-none-manylinux_2_18_x86_64.whl",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-09-06 00:31:43",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "nvidia-nccl-cu12"
}
JSON