{"id":30842716,"url":"https://github.com/nvidia/tilus","last_synced_at":"2025-09-06T21:06:09.232Z","repository":{"id":308759114,"uuid":"1030505915","full_name":"NVIDIA/tilus","owner":"NVIDIA","description":"Tilus is a tile-level kernel programming language with explicit control over shared memory and registers.","archived":false,"fork":false,"pushed_at":"2025-09-04T05:01:22.000Z","size":968,"stargazers_count":324,"open_issues_count":5,"forks_count":8,"subscribers_count":3,"default_branch":"main","last_synced_at":"2025-09-04T06:30:32.255Z","etag":null,"topics":["cuda","kernel","programming","tile"],"latest_commit_sha":null,"homepage":"https://nvidia.github.io/tilus/","language":"Python","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/NVIDIA.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":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-08-01T18:56:09.000Z","updated_at":"2025-09-04T04:51:52.000Z","dependencies_parsed_at":"2025-08-26T23:42:24.033Z","dependency_job_id":"71797a8b-2f6f-494b-824b-e880ba6357b6","html_url":"https://github.com/NVIDIA/tilus","commit_stats":null,"previous_names":["nvidia/tilus"],"tags_count":2,"template":false,"template_full_name":null,"purl":"pkg:github/NVIDIA/tilus","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NVIDIA%2Ftilus","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NVIDIA%2Ftilus/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NVIDIA%2Ftilus/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NVIDIA%2Ftilus/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/NVIDIA","download_url":"https://codeload.github.com/NVIDIA/tilus/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/NVIDIA%2Ftilus/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":273962600,"owners_count":25198601,"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","status":"online","status_checked_at":"2025-09-06T02:00:13.247Z","response_time":2576,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":["cuda","kernel","programming","tile"],"created_at":"2025-09-06T21:06:08.373Z","updated_at":"2025-09-06T21:06:09.207Z","avatar_url":"https://github.com/NVIDIA.png","language":"Python","readme":"# Tilus: A Tile-Level GPU Kernel Programming Language\n[Documentation](https://nvidia.github.io/tilus/index.html) | [Paper](https://arxiv.org/abs/2504.12984)\n\n**Tilus** is a powerful domain-specific language (DSL) for GPU programming that offers:\n\n* **Thread-block-level granularity** with **tensors** as the primary data type.\n* **Explicit control** over shared memory and register tensors (unlike Triton).\n* **Low-precision types** with arbitrary bit-widths (1 to 8 bits).\n\nIt also includes automatic tuning, caching, and a Pythonic interface for ease of use.\n\nTilus is pronounced as tie-lus, /ˈtaɪləs/.\n\n## Getting Started\n\n### Installation\nInstall Tilus using `pip`:\n```\npip install tilus \n```\n\n\u003e [!NOTE]  \n\u003e Tilus depends on `cuda-python`. If your GPU driver is older than **580.65.06**, you will need to install an older version of cuda-python to ensure compatibility.\n\u003e ```\n\u003e pip install tilus \"cuda-python\u003c13\"\n\u003e ```\n\n### Usage\n\nTo get started, refer to the [tutorials](https://nvidia.github.io/tilus/getting-started/tutorials/__init__.html) to learn how to program kernels with Tilus. \n\nYou can also check more [examples](https://github.com/NVIDIA/tilus/tree/main/examples) of using Tilus.\n\nYou can learn more on different topics in the [programming guide](https://nvidia.github.io/tilus/programming-guides/overview.html).\n\n## Research\nThis project is based on the following research paper:\n\n```bibtex\n@article{ding2025tilus,\n  title={Tilus: A Tile-Level GPGPU Programming Language for Low-Precision Computation},\n  author={Ding, Yaoyao and Hou, Bohan and Zhang, Xiao and Lin, Allan and\n    Chen, Tianqi and Hao, Cody Yu and Wang, Yida and Pekhimenko, Gennady},\n  journal={arXiv preprint arXiv:2504.12984},\n  year={2025}\n}\n```\n\n## Acknowledgement\nWe would like to acknowledge the following projects for their influence on Tilus's design and development:\n- **Hidet**: We take Hidet IR as our low-level target and reuse its runtime system.\n- **TVM**: Hidet's initial IR was adopted from TVM, and we also learned a lot from TVM on how to build a compiler.\n- **Triton**: The core idea of defining kernels at a thread-block level and working with tiles was inspired by Triton. \n- **Hexcute**: We adopted the idea of using automatic layout inference to simplify programming from Hexcute.\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnvidia%2Ftilus","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fnvidia%2Ftilus","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnvidia%2Ftilus/lists"}