From f86b0fbfed1583c833b82a091d231839c6954dba Mon Sep 17 00:00:00 2001 From: acostadon Date: Wed, 8 May 2024 13:32:57 -0400 Subject: [PATCH] first copy of general cugraph tutorial. --- .../cugraph/source/tutorials/basic_cugraph.md | 28 +++++++++++++++++++ .../cugraph/source/tutorials/how_to_guides.md | 2 +- 2 files changed, 29 insertions(+), 1 deletion(-) create mode 100644 docs/cugraph/source/tutorials/basic_cugraph.md diff --git a/docs/cugraph/source/tutorials/basic_cugraph.md b/docs/cugraph/source/tutorials/basic_cugraph.md new file mode 100644 index 00000000000..b3486d2d8ba --- /dev/null +++ b/docs/cugraph/source/tutorials/basic_cugraph.md @@ -0,0 +1,28 @@ +## Required hardware/software + +CuGraph is part of [Rapids](https://docs.rapids.ai/user-guide) +It has the following system requirements + * NVIDIA GPU, Volta architecture or later, with [compute capability](https://developer.nvidia.com/cuda-gpus) 7.0+ + * CUDA 11.2, 11.4, 11.5, 11.8, 12.0 or 12.2 + * Python version 3.9, 3.10, or 3.11 + * NetworkX >= version 3.0 (version 3.3 or higher recommended) **This if for use of nx-cuGraph, [see below](#cugraph-using-networkx-code).** + +## Installation +The latest RAPIDS System Requirements documentation is located [here](https://docs.rapids.ai/install#system-req) +This includes several ways to set up for cuGraph +* From Unix + * [Conda](https://docs.rapids.ai/install#wsl-conda) + * [Docker](https://docs.rapids.ai/install#wsl-docker) + * [pip](https://docs.rapids.ai/install#wsl-pip) +* To use RAPIDS in windows you must install [WSL2](https://learn.microsoft.com/en-us/windows/wsl/install) + * [Conda](https://docs.rapids.ai/install#wsl-conda) + * [Docker](https://docs.rapids.ai/install#wsl-docker) + * [pip](https://docs.rapids.ai/install#wsl-pip) +Build From Source + +## CuGraph Using NetworkX Code +While the steps above are required to use the full suite of cuGraph graph analytics, cuGraph is now supported as a NetworkX backend using [nx-cugraph](https://docs.rapids.ai/api/cugraph/nightly/nx_cugraph/nx_cugraph/). +This is much simpler but limits users to the current but growing list of suppored algorithms. + + +## Cugraph API demo \ No newline at end of file diff --git a/docs/cugraph/source/tutorials/how_to_guides.md b/docs/cugraph/source/tutorials/how_to_guides.md index 80be5b4ab5b..998957afea1 100644 --- a/docs/cugraph/source/tutorials/how_to_guides.md +++ b/docs/cugraph/source/tutorials/how_to_guides.md @@ -1,5 +1,5 @@ # How To Guides -- Basic use of cuGraph, on the page +- [Basic use of cuGraph](./basic_cugraph.md) - Property graph with analytic flow - GNN – model building - cuGraph Service – client/server setup and use (ucx)