{"id":13440449,"url":"https://github.com/iree-org/iree-nvgpu","last_synced_at":"2025-03-20T09:33:13.996Z","repository":{"id":153742462,"uuid":"623666488","full_name":"iree-org/iree-nvgpu","owner":"iree-org","description":null,"archived":true,"fork":false,"pushed_at":"2024-03-05T05:27:43.000Z","size":553,"stargazers_count":48,"open_issues_count":17,"forks_count":19,"subscribers_count":6,"default_branch":"main","last_synced_at":"2024-10-28T02:21:02.808Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"MLIR","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/iree-org.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":"AUTHORS","dei":null}},"created_at":"2023-04-04T20:43:44.000Z","updated_at":"2024-09-11T08:21:38.000Z","dependencies_parsed_at":null,"dependency_job_id":"8d1288b1-e4ae-4d7c-a3bd-f8846c4ef87a","html_url":"https://github.com/iree-org/iree-nvgpu","commit_stats":null,"previous_names":["iree-org/iree-nvgpu"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iree-org%2Firee-nvgpu","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iree-org%2Firee-nvgpu/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iree-org%2Firee-nvgpu/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iree-org%2Firee-nvgpu/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/iree-org","download_url":"https://codeload.github.com/iree-org/iree-nvgpu/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":244586050,"owners_count":20476869,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-07-31T03:01:22.875Z","updated_at":"2025-03-20T09:33:13.457Z","avatar_url":"https://github.com/iree-org.png","language":"MLIR","funding_links":[],"categories":["MLIR"],"sub_categories":[],"readme":"# OpenXLA NVIDIA GPU Compiler and Runtime\n\nThis project contains the compiler and runtime plugins enabling specialized\ntargeting of the OpenXLA platform to NVIDIA GPUs. It builds on top of the\ncore IREE toolkit.\n\n## Development setup\n\nThe project can be built either as part of IREE by manually specifying\nplugin paths via `-DIREE_COMPILER_PLUGIN_PATHS`, or for development tailored\nto NVIDIA GPUs specifically, can be built directly:\n\n```\ncmake -GNinja -B build/ -S . \\\n    -DCMAKE_BUILD_TYPE=RelWithDebInfo \\\n    -DIREE_ENABLE_ASSERTIONS=ON \\\n    -DCMAKE_C_COMPILER=clang \\\n    -DCMAKE_CXX_COMPILER=clang++ \\\n    -DIREE_ENABLE_LLD=ON\n\n# Recommended:\n# -DCMAKE_C_COMPILER_LAUNCHER=ccache -DCMAKE_CXX_COMPILER_LAUNCHER=ccache\n```\n\nNote that you will need a check-out of the IREE codebase in `../iree` relative\nto the directory where the `openxla-nvgpu` compiler was checked out. Running\nthe `sync_deps.py` script should bring in all source dependencies at needed\nversions (into the parent directory).\n\nAdditional options for configuring IREE are in the IREE [getting\nstarted](https://openxla.github.io/iree/building-from-source/getting-started/)\nguide for details of how to set this up.\n\n## Installing dependencies\n\nYou must have a CUDA Toolkit installed together with a cuDNN ([see\ninstructions](https://docs.nvidia.com/deeplearning/cudnn/install-guide/index.html#installlinux-tar)).\n\nSee the project settings for options to build without components requiring\nfull dependencies.\n\nOn Linux platform path to `libcudnn.so` should be added to `LD_LIBRARY_PATH`.\n\n```\nexport LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/cuda/lib64\n```\n\n## Running tests\n\nSome of the tests can run only on an Ampere+ devices because they rely on the\n[cuDNN runtime fusion engine](https://docs.nvidia.com/deeplearning/cudnn/developer-guide/index.html#runtime-fusion-engine).\n\nTests that depend on having a device present can be disabled with\n`-DOPENXLA_NVGPU_INCLUDE_DEVICE_TESTS=OFF`.\n\n```\ncmake --build build --target openxla-nvgpu-run-tests\n```\n\n## Project Maintenance\n\nThis section is a work in progress describing various project maintenance\ntasks.\n\n### Pre-requisite: Install openxla-devtools\n\n```\npip install git+https://github.com/openxla/openxla-devtools.git\n```\n\n### Sync all deps to pinned versions\n\n```\nopenxla-workspace sync\n```\n\n### Update IREE to head\n\nThis updates the pinned IREE revision to the HEAD revision at the remote.\n\n```\n# Updates the sync_deps.py metadata.\nopenxla-workspace roll iree\n# Brings all dependencies to pinned versions.\nopenxla-workspace sync\n```\n\n### Full update of all deps\n\nThis updates the pinned revisions of all dependencies. This is presently done\nby updating `openxla-pjrt-plugin` to remote HEAD and deriving the IREE\ndependency from its pin.\n\n```\n# Updates the sync_deps.py metadata.\nopenxla-workspace roll nightly\n# Brings all dependencies to pinned versions.\nopenxla-workspace sync\n```\n\n### Pin current versions of all deps\n\nThis can be done if local, cross project changes have been made and landed.\nIt snapshots the state of all deps as actually checked out and updates\nthe metadata.\n\n```\nopenxla-workspace pin\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Firee-org%2Firee-nvgpu","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Firee-org%2Firee-nvgpu","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Firee-org%2Firee-nvgpu/lists"}