Simulate Moore Threads MUSA and NVIDIA CUDA GPUs in Kubernetes — fake MUSA/MTML/CUDA/NVML user-space libs, mthreads-gmi/nvidia-smi tools, and a device-plugin for HAMi-style scheduling without physical GPUs.
llama.cpp with a DirectX 12 (D3D12) GPU backend: run GGUF LLMs on Windows GPUs without CUDA, ROCm or Vulkan. One zip, no installer. Tested: AMD RDNA 4, Intel Iris Xe, Moore Threads MTT S80.
Easily train a good VC model with voice data <= 10 mins!
This is an index of Moore Threads' game performance / MTT S80游戏性能指数
llama.cpp with a DirectX 11 (D3D11) GPU backend: run GGUF LLMs on older, integrated and low-end Windows GPUs without CUDA, ROCm, Vulkan or DX12. One zip. Tested: AMD RDNA 4, Intel Iris Xe, Moore Threads MTT S80.
musa version for uniad [CVPR 2023 Best Paper Award] Planning-oriented Autonomous Driving
mtgpu-woa —— 在 KVM/QEMU 上运行 Windows 11 ARM64：virtio-gpu + D-Bus 显示、自带 X11 viewer、官方 guest 驱动、摩尔线程 M1000 vfio-pci 直通。中文优先，适配 AIOS（Ubuntu 24.04）。
llama.cpp with an OpenGL 4.3 compute (GLSL) GPU backend: run GGUF LLMs (text, vision, audio) on Windows GPUs without CUDA, ROCm or Vulkan. One zip. Tested: AMD RDNA 4, Intel Iris Xe, Moore Threads MTT S80.
musa version for uniad using OpenMMLab Detection Toolbox and Benchmark
btop++ fork with MUSA (Moore Threads) and other Chinese GPU support, plus an adaptive GPU box layout
To associate your repository with the moore-threads topic, visit your repo's landing page and select "manage topics."