# Cuda project architecture examples

Interactive architecture diagrams of 43 open-source Cuda repositories with a GitDiagram architecture diagram, most stars first. Each diagram shows the project's main components, how they connect, and links every part to its file on GitHub.

## Repositories 1 to 43 of 43

1. [karpathy/llm.c](https://gitdiagram.com/karpathy/llm.c) (31.4k stars): llm.c is a compact LLM pretraining project, centered on GPT-2/GPT-3-family training in C/CUDA, with CPU and PyTorch reference implementations. Markdown: https://gitdiagram.com/karpathy/llm.c.md
2. [nvlabs/instant-ngp](https://gitdiagram.com/nvlabs/instant-ngp) (17.3k stars): Neural Network Engine (based on tiny-cuda-nn) Markdown: https://gitdiagram.com/nvlabs/instant-ngp.md
3. [deepseek-ai/deepep](https://gitdiagram.com/deepseek-ai/deepep) (10.4k stars): DeepEP is a GPU communication library for distributed ML, centered on MoE expert-parallel dispatch and combine. Markdown: https://gitdiagram.com/deepseek-ai/deepep.md
4. [deepseek-ai/deepgemm](https://gitdiagram.com/deepseek-ai/deepgemm) (6.3k stars): API & Interface Layer: Although minimal, the public API exposed via Python calls (e.g., gemm functions and utility functions in deep_gemm/__init__.py) forms the entry point. Markdown: https://gitdiagram.com/deepseek-ai/deepgemm.md
5. [flashinfer-ai/flashinfer](https://gitdiagram.com/flashinfer-ai/flashinfer) (5.2k stars): This project – FlashInfer – is a high-performance kernel library aimed at LLM serving and inference. Markdown: https://gitdiagram.com/flashinfer-ai/flashinfer.md
6. [infatoshi/cuda-course](https://gitdiagram.com/infatoshi/cuda-course) (3.4k stars): Course Modules: Each clearly numbered folder (e.g., “01_Deep_Learning_Ecosystem”, “02_Setup”, “03_C_and_C++_Review”, “04_Gentle_Intro_to_GPUs”, “05_Writing_your_First_Kernels”… Markdown: https://gitdiagram.com/infatoshi/cuda-course.md
7. [hazyresearch/thunderkittens](https://gitdiagram.com/hazyresearch/thunderkittens) (3.3k stars): The ThunderKittens project is a high-performance, CUDA-based framework focused on writing fast deep learning kernels by leveraging tile primitives and fine control of GPU… Markdown: https://gitdiagram.com/hazyresearch/thunderkittens.md
8. [computerhistory/alexnet-source-code](https://gitdiagram.com/computerhistory/alexnet-source-code) (3k stars): This repository is the original 2012 AlexNet source code: a configurable GPU convolutional-neural-network system with ImageNet parameters. Markdown: https://gitdiagram.com/computerhistory/alexnet-source-code.md
9. [mirage-project/mirage](https://gitdiagram.com/mirage-project/mirage) (2.5k stars): Mirage Persistent Kernel (MPK) compiles multi-GPU LLM inference graphs into a persistent fused CUDA “megakernel,” including device computation and communication. Markdown: https://gitdiagram.com/mirage-project/mirage.md
10. [nvidia/cudalibrarysamples](https://gitdiagram.com/nvidia/cudalibrarysamples) (2.4k stars): This project is an extensive repository of CUDA library samples that demonstrates a wide array of GPU-accelerated libraries for different computational domains. Markdown: https://gitdiagram.com/nvidia/cudalibrarysamples.md
11. [siboehm/sgemm_cuda](https://gitdiagram.com/siboehm/sgemm_cuda) (1.3k stars) Markdown: https://gitdiagram.com/siboehm/sgemm_cuda.md
12. [k2-fsa/k2](https://gitdiagram.com/k2-fsa/k2) (1.3k stars) Markdown: https://gitdiagram.com/k2-fsa/k2.md
13. [packtpublishing/learn-cuda-programming](https://gitdiagram.com/packtpublishing/learn-cuda-programming) (1.2k stars) Markdown: https://gitdiagram.com/packtpublishing/learn-cuda-programming.md
14. [tongkaio/cuda_kernel_samples](https://gitdiagram.com/tongkaio/cuda_kernel_samples) (1.1k stars) Markdown: https://gitdiagram.com/tongkaio/cuda_kernel_samples.md
15. [woct0rdho/sageattention](https://gitdiagram.com/woct0rdho/sageattention) (1.1k stars) Markdown: https://gitdiagram.com/woct0rdho/sageattention.md
16. [nvidia/nvbench](https://gitdiagram.com/nvidia/nvbench) (839 stars) Markdown: https://gitdiagram.com/nvidia/nvbench.md
17. [nvidia/cuopt](https://gitdiagram.com/nvidia/cuopt) (788 stars) Markdown: https://gitdiagram.com/nvidia/cuopt.md
18. [brucefan1983/gpumd](https://gitdiagram.com/brucefan1983/gpumd) (742 stars) Markdown: https://gitdiagram.com/brucefan1983/gpumd.md
19. [clu0/unet.cu](https://gitdiagram.com/clu0/unet.cu) (666 stars) Markdown: https://gitdiagram.com/clu0/unet.cu.md
20. [daviddiazguerra/gpurir](https://gitdiagram.com/daviddiazguerra/gpurir) (611 stars) Markdown: https://gitdiagram.com/daviddiazguerra/gpurir.md
21. [pranjalssh/fast.cu](https://gitdiagram.com/pranjalssh/fast.cu) (561 stars) Markdown: https://gitdiagram.com/pranjalssh/fast.cu.md
22. [xuezhemax/megalodon](https://gitdiagram.com/xuezhemax/megalodon) (528 stars) Markdown: https://gitdiagram.com/xuezhemax/megalodon.md
23. [alicevision/popsift](https://gitdiagram.com/alicevision/popsift) (491 stars) Markdown: https://gitdiagram.com/alicevision/popsift.md
24. [mw00/project-maya](https://gitdiagram.com/mw00/project-maya) (484 stars) Markdown: https://gitdiagram.com/mw00/project-maya.md
25. [moorethreads/litegs](https://gitdiagram.com/moorethreads/litegs) (417 stars) Markdown: https://gitdiagram.com/moorethreads/litegs.md
26. [mit-han-lab/quest](https://gitdiagram.com/mit-han-lab/quest) (410 stars) Markdown: https://gitdiagram.com/mit-han-lab/quest.md
27. [acdslab/mppi-generic](https://gitdiagram.com/acdslab/mppi-generic) (263 stars) Markdown: https://gitdiagram.com/acdslab/mppi-generic.md
28. [dots-infra/ultraep](https://gitdiagram.com/dots-infra/ultraep) (251 stars) Markdown: https://gitdiagram.com/dots-infra/ultraep.md
29. [flamegpu/flamegpu2](https://gitdiagram.com/flamegpu/flamegpu2) (145 stars) Markdown: https://gitdiagram.com/flamegpu/flamegpu2.md
30. [bchao1/textured_gaussians](https://gitdiagram.com/bchao1/textured_gaussians) (109 stars) Markdown: https://gitdiagram.com/bchao1/textured_gaussians.md
31. [microsoft/tilefusion](https://gitdiagram.com/microsoft/tilefusion) (106 stars) Markdown: https://gitdiagram.com/microsoft/tilefusion.md
32. [rocm/rccl-tests](https://gitdiagram.com/rocm/rccl-tests) (88 stars) Markdown: https://gitdiagram.com/rocm/rccl-tests.md
33. [owensgroup/slabhash](https://gitdiagram.com/owensgroup/slabhash) (77 stars) Markdown: https://gitdiagram.com/owensgroup/slabhash.md
34. [jvhs0706/zkllm-ccs2024](https://gitdiagram.com/jvhs0706/zkllm-ccs2024) (58 stars) Markdown: https://gitdiagram.com/jvhs0706/zkllm-ccs2024.md
35. [owensgroup/gpubtree](https://gitdiagram.com/owensgroup/gpubtree) (58 stars) Markdown: https://gitdiagram.com/owensgroup/gpubtree.md
36. [vaktibabat/cudacracker](https://gitdiagram.com/vaktibabat/cudacracker) (47 stars) Markdown: https://gitdiagram.com/vaktibabat/cudacracker.md
37. [rox906/tcfft](https://gitdiagram.com/rox906/tcfft) (43 stars) Markdown: https://gitdiagram.com/rox906/tcfft.md
38. [higherorderco/hvm2](https://gitdiagram.com/higherorderco/hvm2) (41 stars) Markdown: https://gitdiagram.com/higherorderco/hvm2.md
39. [yigongqin/cupf](https://gitdiagram.com/yigongqin/cupf) (33 stars) Markdown: https://gitdiagram.com/yigongqin/cupf.md
40. [openhero/im2col](https://gitdiagram.com/openhero/im2col) (30 stars) Markdown: https://gitdiagram.com/openhero/im2col.md
41. [fishaudio/fish-scales-ops](https://gitdiagram.com/fishaudio/fish-scales-ops) (28 stars) Markdown: https://gitdiagram.com/fishaudio/fish-scales-ops.md
42. [ntnu-hpc-lab/bat](https://gitdiagram.com/ntnu-hpc-lab/bat) (19 stars) Markdown: https://gitdiagram.com/ntnu-hpc-lab/bat.md
43. [andreluizbvs/voxarmps](https://gitdiagram.com/andreluizbvs/voxarmps) (11 stars) Markdown: https://gitdiagram.com/andreluizbvs/voxarmps.md

More hubs: https://gitdiagram.com/architecture-examples

Counted over GitDiagram's stored diagrams of public repositories, from the names their components and groups carry and the files they link to; the rules are the ones in the [2026 open-source architecture report](https://gitdiagram.com/open-source-architecture-2026). Stars are GitHub's when the diagram was last made. Rebuilt daily: hubs grow as more repositories' languages are known.