USE Flag: cuda

Provenance Summary

Authoritative Scope: global_defined | Distinct Packages: 132 | Repository Package Instances: 222

Search

Manage

Global / USE_EXPAND Definitions

Repository Source Kind Source File Description
gentoo use.desc profiles/use.desc Enable NVIDIA CUDA support (computation on GPU)

Package-Scoped Descriptions (use.local.desc / metadata.xml)

Repository Package Source Kind Source File Description
Miezhiko media-libs/opencv metadata.xml media-libs/opencv/metadata.xml Enable NVIDIA Cuda computations support (Experimental!)
baar-dau media-gfx/blender metadata.xml media-gfx/blender/metadata.xml Build cycles renderer with nVidia CUDA support.
bennypowers dev-python/comfy-aimdo metadata.xml dev-python/comfy-aimdo/metadata.xml Build the NVIDIA backend (aimdo.so)
bennypowers media-gfx/comfyui metadata.xml media-gfx/comfyui/metadata.xml Enable NVIDIA CUDA support via sci-ml/pytorch and build the comfy-aimdo CUDA backend
bennypowers sci-ml/colibri metadata.xml sci-ml/colibri/metadata.xml Build with NVIDIA CUDA GPU offloading for resident expert tensors
bentoo media-gfx/blender metadata.xml media-gfx/blender/metadata.xml Build cycles renderer with nVidia CUDA support.
bentoo media-gfx/blender use.local.desc profiles/use.local.desc Build cycles renderer with nVidia CUDA support.
bentoo media-gfx/openvdb metadata.xml media-gfx/openvdb/metadata.xml Enable support for CUDA in NanoVDB
bentoo media-libs/opencv metadata.xml media-libs/opencv/metadata.xml Enable NVIDIA Cuda computations support (Experimental!)
bentoo media-libs/opencv use.local.desc profiles/use.local.desc Enable NVIDIA Cuda computations support (Experimental!)
bentoo sci-libs/dlib metadata.xml sci-libs/dlib/metadata.xml Enable support for CUDA for Deep Neural Networks (cuDNN) on GPU (experimental)
bentoo sci-ml/koboldcpp metadata.xml sci-ml/koboldcpp/metadata.xml Build the CUDA/cuBLAS backend (koboldcpp_cublas) for NVIDIA GPUs, for every architecture the installed toolkit supports
bentoo sci-ml/koboldcpp use.local.desc profiles/use.local.desc Build the CUDA/cuBLAS backend (koboldcpp_cublas) for NVIDIA GPUs, for every architecture the installed toolkit supports
bentoo sci-ml/ollama-bin metadata.xml sci-ml/ollama-bin/metadata.xml Enable NVIDIA CUDA GPU acceleration support. Requires compatible NVIDIA GPU (compute capability 6.0+, Pascal architecture or newer) and nvidia-cuda-toolkit. Significantly improves inference performance for large models.
bentoo sci-ml/ollama-bin use.local.desc profiles/use.local.desc Enable NVIDIA CUDA GPU acceleration support. Requires compatible NVIDIA GPU (compute capability 6.0+, Pascal architecture or newer) and nvidia-cuda-toolkit. Significantly improves inference performance for large models.
bentoo sci-ml/sherpa-onnx metadata.xml sci-ml/sherpa-onnx/metadata.xml Build the NVIDIA CUDA execution provider (requires the CUDA Toolkit; the nvcc host compiler is resolved from the installed toolkit, not hardcoded)
bentoo sci-ml/sherpa-onnx use.local.desc profiles/use.local.desc Build the NVIDIA CUDA execution provider (requires the CUDA Toolkit; the nvcc host compiler is resolved from the installed toolkit, not hardcoded)
bentoo sci-ml/torchaudio metadata.xml sci-ml/torchaudio/metadata.xml Build the CUDA kernels (RNN-T loss, forced alignment, IIR filter) and the CUDA CTC decoder
bentoo sci-ml/torchcodec metadata.xml sci-ml/torchcodec/metadata.xml Build the NVIDIA GPU decoders (NVDEC through FFmpeg, nvJPEG)
blshkv dev-python/nvidia-apex metadata.xml dev-python/nvidia-apex/metadata.xml Add support for CUDA processing
bobwya net-misc/gminer-bin metadata.xml net-misc/gminer-bin/metadata.xml Support Nvidia graphics cards
bobwya net-misc/lolminer-bin metadata.xml net-misc/lolminer-bin/metadata.xml Support Nvidia graphics cards
bobwya net-misc/srbminer-multi-bin metadata.xml net-misc/srbminer-multi-bin/metadata.xml Support Nvidia graphics cards
bobwya net-misc/xmrig metadata.xml net-misc/xmrig/metadata.xml Build cuda plugin
bobwya net-misc/xmrigcc metadata.xml net-misc/xmrigcc/metadata.xml Build cuda plugin
dev1990-overlay app-misc/ollama metadata.xml app-misc/ollama/metadata.xml Enable NVIDIA CUDA support
ex_repo dev-cpp/kokkos metadata.xml dev-cpp/kokkos/metadata.xml Enable the CUDA execution space. Any Kokkos application compiled for CUDA embeds CUDA code via template metaprogramming. Thus, the whole application must be built with a CUDA-capable compiler. (At the moment, the only such compilers are NVIDIA’s NVCC and Clang 10.0+)
gentoo media-gfx/blender metadata.xml media-gfx/blender/metadata.xml Build cycles renderer with nVidia CUDA support.
gentoo media-gfx/blender use.local.desc profiles/use.local.desc Build cycles renderer with nVidia CUDA support.
gentoo media-gfx/openvdb metadata.xml media-gfx/openvdb/metadata.xml Enable support for CUDA in NanoVDB
gentoo media-gfx/openvdb use.local.desc profiles/use.local.desc Enable support for CUDA in NanoVDB
gentoo media-libs/opencv metadata.xml media-libs/opencv/metadata.xml Enable NVIDIA Cuda computations support (Experimental!)
gentoo media-libs/opencv use.local.desc profiles/use.local.desc Enable NVIDIA Cuda computations support (Experimental!)
gentoo net-misc/sunshine metadata.xml net-misc/sunshine/metadata.xml Enable accelerated video encoding on NVIDIA hardware
gentoo net-misc/sunshine use.local.desc profiles/use.local.desc Enable accelerated video encoding on NVIDIA hardware
gentoo sci-chemistry/gromacs metadata.xml sci-chemistry/gromacs/metadata.xml Enable cuda non-bonded kernels
gentoo sci-chemistry/gromacs use.local.desc profiles/use.local.desc Enable cuda non-bonded kernels
gentoo sci-libs/clblast metadata.xml sci-libs/clblast/metadata.xml Build with support for cuda instead of opencl (beta!)
gentoo sci-libs/clblast use.local.desc profiles/use.local.desc Build with support for cuda instead of opencl (beta!)
gentoo sci-libs/dlib metadata.xml sci-libs/dlib/metadata.xml Enable support for CUDA for Deep Neural Networks (cuDNN) on GPU (experimental)
gentoo sci-libs/dlib use.local.desc profiles/use.local.desc Enable support for CUDA for Deep Neural Networks (cuDNN) on GPU (experimental)
gentoo sci-libs/ginkgo metadata.xml sci-libs/ginkgo/metadata.xml Add support for cuda assimp (<pkg>dev-util/nvidia-cuda-toolkit</pkg>)
gentoo sci-libs/ginkgo use.local.desc profiles/use.local.desc Add support for cuda assimp (dev-util/nvidia-cuda-toolkit)
gentoo sys-apps/hwloc metadata.xml sys-apps/hwloc/metadata.xml Enable CUDA device discovery using libcudart
gentoo sys-apps/hwloc use.local.desc profiles/use.local.desc Enable CUDA device discovery using libcudart
gentooplusplus sci-ml/llama-cpp metadata.xml sci-ml/llama-cpp/metadata.xml ggml: use CUDA ; in order for it to compile also tag .... "cpu" has to be selected, don't ask why. Use CMAKE_EXTRA_CACHE_FILE env variable (check https://gitweb.gentoo.org/repo/gentoo.git/tree/eclass/cmake.eclass ) to specify next variables: CMAKE_CUDA_ARCHITECTURES ( use next command to find native architecture: `nvidia-smi --query-gpu=compute_cap --format=csv | tail -n 1 | sed -e 's/\.//g'` ) ; GGML_CUDA_PEER_MAX_BATCH_SIZE ggml: max. batch size for using peer access, default: 128
ha-bleeding-edge sci-libs/tensorflow metadata.xml sci-libs/tensorflow/metadata.xml Pull archive/v2.8.3-1.tar.gz for cuda support
hep-forge sci-libs/einspline metadata.xml sci-libs/einspline/metadata.xml Build with cuda
junkdrawer dev-libs/pocl metadata.xml dev-libs/pocl/metadata.xml Enable the CUDA backend for NVIDIA GPUs
kzd media-gfx/blender metadata.xml media-gfx/blender/metadata.xml Build cycles renderer with nVidia CUDA support.
kzd net-misc/sunshine metadata.xml net-misc/sunshine/metadata.xml Enable accelerated video encoding on NVIDIA hardware
local-ai localai-backend/tinygrad metadata.xml localai-backend/tinygrad/metadata.xml Run on NVIDIA GPUs through tinygrad's CUDA device
local-ai sci-ml/bitsandbytes metadata.xml sci-ml/bitsandbytes/metadata.xml Build the CUDA quantization kernels (COMPUTE_BACKEND=cuda)
local-ai sci-ml/torchaudio metadata.xml sci-ml/torchaudio/metadata.xml GPU acceleration via CUDA
local-ai sci-ml/torchcodec metadata.xml sci-ml/torchcodec/metadata.xml NVDEC video decoding and nvJPEG via CUDA
mim media-libs/opencv metadata.xml media-libs/opencv/metadata.xml Enable NVIDIA Cuda computations support (Experimental!)
mim sci-libs/dlib metadata.xml sci-libs/dlib/metadata.xml Enable support for CUDA for Deep Neural Networks (cuDNN) on GPU (experimental)
mim sys-apps/hwloc metadata.xml sys-apps/hwloc/metadata.xml Enable CUDA device discovery using libcudart
pf4public sci-misc/llama-cpp metadata.xml sci-misc/llama-cpp/metadata.xml Enable the NVIDIA CUDA GPU backend
pf4public sci-misc/stable-diffusion-cpp metadata.xml sci-misc/stable-diffusion-cpp/metadata.xml Enable the NVIDIA CUDA GPU backend
pf4public sci-misc/strata metadata.xml sci-misc/strata/metadata.xml Build the engine for NVIDIA GPUs (CUDA backend, sm_75 through sm_120: RTX 20/30/40/50)
pf4public sci-misc/whisper-cpp metadata.xml sci-misc/whisper-cpp/metadata.xml Enable the NVIDIA CUDA backend (GGML_CUDA)
pf4public sci-ml/ggml metadata.xml sci-ml/ggml/metadata.xml Enable the NVIDIA CUDA GPU backend
pg_overlay media-libs/opencv metadata.xml media-libs/opencv/metadata.xml Enable NVIDIA Cuda computations support (Experimental!)
pg_overlay sci-mathematics/mathematica metadata.xml sci-mathematics/mathematica/metadata.xml Install with cuda support
pingwho-overlay sci-ml/ollama-bin metadata.xml sci-ml/ollama-bin/metadata.xml Support for Nvidia CUDA compatible cards
pkalin net-analyzer/suricata metadata.xml net-analyzer/suricata/metadata.xml Enable NVIDIA Cuda computations support
pypi-sci dev-python/cupy metadata.xml dev-python/cupy/metadata.xml Enable support for nVidia CUDA
ruizhi-overlay dev-python/comfy-kitchen metadata.xml dev-python/comfy-kitchen/metadata.xml Build NVIDIA CUDA kernels
ruizhi-overlay media-gfx/comfyui metadata.xml media-gfx/comfyui/metadata.xml Use CUDA-enabled PyTorch, torchvision and torchaudio operators; CPU is the default when neither GPU flag is enabled.
ruizhi-overlay sci-ml/strata metadata.xml sci-ml/strata/metadata.xml Build the engine for NVIDIA GPUs with the CUDA toolkit
ruizhi-overlay sci-ml/torchaudio metadata.xml sci-ml/torchaudio/metadata.xml Build CUDA audio operators to match a CUDA-enabled PyTorch.
sage-on-gentoo sci-libs/cholmod metadata.xml sci-libs/cholmod/metadata.xml Use nvidia cuda toolkit for speeding up computations
science dev-libs/boost-compute metadata.xml dev-libs/boost-compute/metadata.xml build cuda examples and tests
science sci-biology/fsl metadata.xml sci-biology/fsl/metadata.xml Add support for CUDA
science sci-biology/megahit metadata.xml sci-biology/megahit/metadata.xml Enable NVIDIA GTX680 (4G memory) and Tesla K40c (12G memory) with CUDA 5.5, 6.0 and 6.5
science sci-chemistry/relion metadata.xml sci-chemistry/relion/metadata.xml Enable CUDA support
science sci-libs/arrayfire metadata.xml sci-libs/arrayfire/metadata.xml Build CUDA backend.
science sys-devel/DPC++ metadata.xml sys-devel/DPC++/metadata.xml use the cuda backend
stuff app-accessibility/whisper-cpp metadata.xml app-accessibility/whisper-cpp/metadata.xml Enable NVIDIA CUDA GPU backend
stuff dev-python/cupy metadata.xml dev-python/cupy/metadata.xml Enable support for nVidia CUDA
stuff dev-python/sasmodels metadata.xml dev-python/sasmodels/metadata.xml Enable CUDA GPU-accelerated model evaluation via <pkg>dev-python/pycuda</pkg>
stuff media-gfx/comfyui metadata.xml media-gfx/comfyui/metadata.xml Build the torch backend (<pkg>sci-ml/caffe2</pkg>) with CUDA/NVIDIA acceleration; also selects the CUDA wheels of <pkg>dev-python/comfy-aimdo-bin</pkg> and <pkg>dev-python/comfy-kitchen-bin</pkg> (otherwise their non-CUDA wheels)
stuff media-libs/opencv metadata.xml media-libs/opencv/metadata.xml Enable NVIDIA Cuda computations support (Experimental!)
stuff sci-libs/cholmod metadata.xml sci-libs/cholmod/metadata.xml Use nvidia cuda toolkit for speeding up computations
stuff sci-libs/onnxruntime metadata.xml sci-libs/onnxruntime/metadata.xml Build the NVIDIA CUDA execution provider
stuff sci-ml/ollama metadata.xml sci-ml/ollama/metadata.xml Build the CUDA (cuda_v13) llama-server GPU backend via <pkg>dev-util/nvidia-cuda-toolkit</pkg>
stuff sci-ml/sherpa-onnx metadata.xml sci-ml/sherpa-onnx/metadata.xml Build the NVIDIA CUDA backend (requires CUDA Toolkit and a compiler version supported by nvcc)
stuff sci-ml/torchcodec metadata.xml sci-ml/torchcodec/metadata.xml Build the NVIDIA CUDA decoding/encoding backend (libnpp + libnvrtc)
touchfish-os sci-physics/lammps metadata.xml sci-physics/lammps/metadata.xml Enable cuda gpu computing support
waebbl media-libs/opensubdiv metadata.xml media-libs/opensubdiv/metadata.xml Enable NVIDIA CUDA Toolkit support through <pkg>dev-util/nvidia-cuda-toolkit</pkg>
waebbl media-libs/opensubdiv use.local.desc profiles/use.local.desc Enable NVIDIA CUDA Toolkit support through dev-util/nvidia-cuda-toolkit
zGentoo media-libs/opencv metadata.xml media-libs/opencv/metadata.xml Enable NVIDIA Cuda computations support (Experimental!)

Package Usage Occurrences

Repository Package Observations Versions Local Description (use.local.desc) Metadata Description (metadata.xml)
FireBurn sci-misc/llama-cpp IUSE 1.0.0, 9999 - -
FireBurn sci-misc/stable-diffusion-cpp IUSE 1106, 9999 - -
Miezhiko media-libs/opencv IUSE + REQUIRED_USE 5.0.0 - Enable NVIDIA Cuda computations support (Experimental!)
baar-dau media-gfx/blender IUSE + REQUIRED_USE 5.1.2, 5.2.0 - Build cycles renderer with nVidia CUDA support.
baar-dau sci-libs/ceres-solver IUSE 2.2.0 - -
bennypowers dev-python/comfy-aimdo IUSE 0.5.6, 0.6.0 - Build the NVIDIA backend (aimdo.so)
bennypowers dev-python/comfy-kitchen IUSE 0.2.39, 0.2.40, 0.2.41, 0.2.42 - -
bennypowers media-gfx/comfyui IUSE 0.42.1, 0.43.0, 0.44.0, 0.45.0, 0.46.0 - Enable NVIDIA CUDA support via sci-ml/pytorch and build the comfy-aimdo CUDA backend
bennypowers sci-ml/colibri IUSE + REQUIRED_USE 2.1.0 - Build with NVIDIA CUDA GPU offloading for resident expert tensors
bentoo dev-cpp/eigen IUSE 3.4.1 - -
bentoo dev-util/goose IUSE 1.59.1, 1.60.2 - -
bentoo llvm-runtimes/openmp IUSE 23.1.6 - -
bentoo media-gfx/blender IUSE + REQUIRED_USE 5.3.0 Build cycles renderer with nVidia CUDA support. Build cycles renderer with nVidia CUDA support.
bentoo media-gfx/blender-bin IUSE 5.3.0 - -
bentoo media-gfx/openvdb IUSE 14.0.0 - Enable support for CUDA in NanoVDB
bentoo media-libs/oidn IUSE 2.5.1 - -
bentoo media-libs/opencv IUSE + REQUIRED_USE 5.1.0 Enable NVIDIA Cuda computations support (Experimental!) Enable NVIDIA Cuda computations support (Experimental!)
bentoo media-libs/openimageio IUSE 3.2.2.0 - -
bentoo media-libs/opensubdiv IUSE 3.8.0 - -
bentoo sci-libs/dlib IUSE 20.1 - Enable support for CUDA for Deep Neural Networks (cuDNN) on GPU (experimental)
bentoo sci-ml/ik_llama-cpp IUSE 0_pre20261112, 0_pre20261121 - -
bentoo sci-ml/koboldcpp IUSE 1.124.2 Build the CUDA/cuBLAS backend (koboldcpp_cublas) for NVIDIA GPUs, for every architecture the installed toolkit supports Build the CUDA/cuBLAS backend (koboldcpp_cublas) for NVIDIA GPUs, for every architecture the installed toolkit supports
bentoo sci-ml/llama-cpp IUSE 0.9.0, 1.0.0 - -
bentoo sci-ml/ollama-bin IUSE + REQUIRED_USE 0.40.16, 0.40.17 Enable NVIDIA CUDA GPU acceleration support. Requires compatible NVIDIA GPU (compute capability 6.0+, Pascal architecture or newer) and nvidia-cuda-toolkit. Significantly improves inference performance for large models. Enable NVIDIA CUDA GPU acceleration support. Requires compatible NVIDIA GPU (compute capability 6.0+, Pascal architecture or newer) and nvidia-cuda-toolkit. Significantly improves inference performance for large models.
bentoo sci-ml/sherpa-onnx IUSE 2.0.1, 2.0.2 Build the NVIDIA CUDA execution provider (requires the CUDA Toolkit; the nvcc host compiler is resolved from the installed toolkit, not hardcoded) Build the NVIDIA CUDA execution provider (requires the CUDA Toolkit; the nvcc host compiler is resolved from the installed toolkit, not hardcoded)
bentoo sci-ml/stable-diffusion-cpp IUSE 0_pre782 - -
bentoo sci-ml/torchaudio IUSE 2.11.1 - Build the CUDA kernels (RNN-T loss, forced alignment, IIR filter) and the CUDA CTC decoder
bentoo sci-ml/torchcodec IUSE 0.17.0 - Build the NVIDIA GPU decoders (NVDEC through FFmpeg, nvJPEG)
bentoo sci-ml/whisper-cpp IUSE 1.9.5 - -
blshkv dev-libs/caffe IUSE 1.0 - -
blshkv dev-python/nvidia-apex IUSE 23.08, 24.04.01 - Add support for CUDA processing
blshkv media-gfx/ideepcolor IUSE 9999 - -
blshkv sci-libs/tensorflow IUSE 2.15.0 - -
bobwya net-misc/gminer-bin IUSE 3.43 - Support Nvidia graphics cards
bobwya net-misc/lolminer-bin IUSE 1.84 - Support Nvidia graphics cards
bobwya net-misc/srbminer-multi-bin IUSE 2.4.2, 2.4.3, 2.4.4, 2.4.5, 2.4.7, 2.4.8, 2.4.9 - Support Nvidia graphics cards
bobwya net-misc/xmrig IUSE 6.21.1, 9999 - Build cuda plugin
bobwya net-misc/xmrigcc IUSE 3.4.0, 9999 - Build cuda plugin
booboo app-accessibility/whisper-cpp IUSE 1.9.2, 1.9.4 - -
booboo media-gfx/stable-diffusion-cpp IUSE 9999 - -
booboo sci-misc/llama-cpp IUSE 0.7.0, 0.8.0, 0.9.0, 1.0.0, 9999 - -
cy1der media-gfx/f3d IUSE 4.0.1 - -
cy1der sci-libs/vtk IUSE + REQUIRED_USE 9.7.2 - -
dev1990-overlay app-misc/ollama IUSE 0.5.7 - Enable NVIDIA CUDA support
ex_repo dev-cpp/kokkos IUSE 4.1.0 - Enable the CUDA execution space. Any Kokkos application compiled for CUDA embeds CUDA code via template metaprogramming. Thus, the whole application must be built with a CUDA-capable compiler. (At the moment, the only such compilers are NVIDIA’s NVCC and Clang 10.0+)
gentoo acct-user/ollama IUSE 0 - -
gentoo app-containers/docker IUSE 28.4.0, 29.10.0, 29.7.2, 29.8.0, 29.8.1, 29.9.0, 29.9.3 - -
gentoo dev-cpp/eigen IUSE 3.4.1, 3.4.9999, 5.0.1, 5.0.9999, 9999 - -
gentoo dev-cpp/kokkos IUSE 4.6.2, 4.7.4, 5.0.2, 5.1.1, 9999 - -
gentoo dev-libs/pocl IUSE 7.0, 7.1 - -
gentoo dev-libs/starpu IUSE 1.3.9 - -
gentoo dev-util/bear IUSE 3.1.6 - -
gentoo llvm-runtimes/openmp IUSE 23.1.6, 24.0.0.9999, 24.0.0_pre20261115, 24.0.0_pre20261121 - -
gentoo media-gfx/blender IUSE + REQUIRED_USE 4.5.5, 4.5.9999, 5.0.0, 5.0.9999, 9999 Build cycles renderer with nVidia CUDA support. Build cycles renderer with nVidia CUDA support.
gentoo media-gfx/blender-bin IUSE 4.2.23, 4.3.2, 4.4.3, 4.5.15, 5.0.1, 5.1.2, 5.2.3, 5.3.0, 9999 - -
gentoo media-gfx/openvdb IUSE 12.0.1 Enable support for CUDA in NanoVDB Enable support for CUDA in NanoVDB
gentoo media-libs/oidn IUSE 2.3.2, 2.3.3 - -
gentoo media-libs/opencv IUSE + REQUIRED_USE 4.11.0, 4.12.0, 4.14.0 Enable NVIDIA Cuda computations support (Experimental!) Enable NVIDIA Cuda computations support (Experimental!)
gentoo media-libs/openimageio IUSE 3.0.11.0, 3.0.6.1, 3.0.8.1, 3.1.7.0, 9999 - -
gentoo media-libs/opensubdiv IUSE 3.6.0, 3.6.1 - -
gentoo net-analyzer/suricata IUSE 7.0.15 - -
gentoo net-misc/sunshine IUSE 2026.516.143833, 2026.914.233613, 9999 Enable accelerated video encoding on NVIDIA hardware Enable accelerated video encoding on NVIDIA hardware
gentoo sci-chemistry/gromacs IUSE + REQUIRED_USE 2023.5, 2024.6, 2025.3, 2025.4, 2025.5, 2026.2, 2026.3, 2026.4, 2026.9999, 9999 Enable cuda non-bonded kernels Enable cuda non-bonded kernels
gentoo sci-chemistry/vmd IUSE 1.9.4_alpha57 - -
gentoo sci-libs/ceres-solver IUSE 2.2.0 - -
gentoo sci-libs/cholmod IUSE 3.0.13, 3.0.14, 5.3.4 - -
gentoo sci-libs/clblast IUSE + REQUIRED_USE 1.5.2, 1.5.3, 1.6.3 Build with support for cuda instead of opencl (beta!) Build with support for cuda instead of opencl (beta!)
gentoo sci-libs/dlib IUSE 20.0, 20.0.1 Enable support for CUDA for Deep Neural Networks (cuDNN) on GPU (experimental) Enable support for CUDA for Deep Neural Networks (cuDNN) on GPU (experimental)
gentoo sci-libs/flann IUSE 1.9.2_p20221027 - -
gentoo sci-libs/ginkgo IUSE 1.10.0, 1.11.0, 1.4.0, 1.6.0, 1.8.0 Add support for cuda assimp (dev-util/nvidia-cuda-toolkit) Add support for cuda assimp (<pkg>dev-util/nvidia-cuda-toolkit</pkg>)
gentoo sci-libs/pastix IUSE 6.0.3 - -
gentoo sci-libs/pcl IUSE 1.14.1, 1.15.1, 9999 - -
gentoo sci-libs/spqr IUSE 4.3.6 - -
gentoo sci-libs/suitesparse IUSE 5.4.0, 7.12.2 - -
gentoo sci-libs/suitesparseconfig IUSE 7.12.2 - -
gentoo sci-libs/trilinos IUSE 16.2.2, 17.1.2, 17.2.1 - -
gentoo sci-libs/vtk IUSE + REQUIRED_USE 9.5.2, 9.6.2, 9999 - -
gentoo sci-mathematics/mathematica IUSE 13.0.1, 13.1.0, 13.2.1, 13.3.1, 14.0.0, 14.1.0, 14.2.1, 14.3.0, 15.0.1 - -
gentoo sci-mathematics/petsc IUSE 3.23.3, 3.23.4, 3.24.5, 3.25.1 - -
gentoo sci-misc/boinc IUSE 7.24.1 - -
gentoo sci-ml/caffe2 IUSE + REQUIRED_USE 2.12.0 - -
gentoo sci-ml/ggml IUSE 1.0.0 - -
gentoo sci-ml/gloo IUSE 2023.12.03, 2025.06.04, 2025.12.02, 2026.06.10 - -
gentoo sci-ml/kineto IUSE 0.4.0_p20260805 - -
gentoo sci-ml/pytorch IUSE + REQUIRED_USE 2.13.0, 2.14.1, 2.15.0 - -
gentoo sci-ml/tensorpipe IUSE 2025.11.05 - -
gentoo sci-ml/torchvision IUSE + REQUIRED_USE 0.27.0, 0.28.0, 0.29.1, 0.29.2 - -
gentoo sci-physics/root IUSE + REQUIRED_USE 6.38.04, 6.40.04, 6.40.9999, 9999 - -
gentoo sys-apps/hwloc IUSE 2.11.2, 2.12.2 Enable CUDA device discovery using libcudart Enable CUDA device discovery using libcudart
gentoo sys-block/libfabric IUSE 1.11.2, 2.3.1 - -
gentoo sys-cluster/openmpi IUSE 4.1.6, 5.0.6, 5.0.8, 5.0.9 - -
gentoo x11-wm/xpra IUSE + REQUIRED_USE 6.2.5, 6.4.4, 6.9999, 9999 - -
gentoo-zh sci-ml/llama-cpp IUSE 1.0.0, 9999 - -
gentooplusplus sci-ml/llama-cpp IUSE + REQUIRED_USE 9999 - ggml: use CUDA ; in order for it to compile also tag .... "cpu" has to be selected, don't ask why. Use CMAKE_EXTRA_CACHE_FILE env variable (check https://gitweb.gentoo.org/repo/gentoo.git/tree/eclass/cmake.eclass ) to specify next variables: CMAKE_CUDA_ARCHITECTURES ( use next command to find native architecture: `nvidia-smi --query-gpu=compute_cap --format=csv | tail -n 1 | sed -e 's/\.//g'` ) ; GGML_CUDA_PEER_MAX_BATCH_SIZE ggml: max. batch size for using peer access, default: 128
guru app-accessibility/whisper-cpp IUSE 1.8.4, 1.9.4, 9999 - -
guru sci-misc/ik_llama-cpp IUSE 9999 - -
guru sci-misc/llama-cpp IUSE 0.7.0, 0.8.0, 9999 - -
guru sci-misc/stable-diffusion-cpp IUSE 9999 - -
ha-bleeding-edge sci-libs/tensorflow IUSE 2.5.0 - Pull archive/v2.8.3-1.tar.gz for cuda support
hep-forge sci-libs/einspline IUSE 0.9.2 - Build with cuda
hossie media-gfx/blender-bin IUSE 4.1.1, 4.2.0, 5.1.1 - -
hossie sci-misc/llama-cpp IUSE 0_pre10453, 0_pre9747, 0_pre9941 - -
inode64-overlay sci-misc/llama-cpp IUSE 9999 - -
junkdrawer dev-libs/pocl IUSE 9999 - Enable the CUDA backend for NVIDIA GPUs
kzd media-gfx/blender IUSE + REQUIRED_USE 5.1.0 - Build cycles renderer with nVidia CUDA support.
kzd media-libs/opensubdiv IUSE 3.6.1 - -
kzd net-misc/sunshine IUSE 2026.516.143833 - Enable accelerated video encoding on NVIDIA hardware
lamdness acct-user/ollama IUSE 3 - -
lamdness app-misc/ollama-bin IUSE 0.31.1, 0.32.5, 0.34.4, 0.40.0, 0.40.16, 0.40.6 - -
lastrodamo media-gfx/luxcorerender IUSE 9999 - -
local-ai localai-backend/audio-cpp IUSE + REQUIRED_USE 5.0.1 - -
local-ai localai-backend/tinygrad IUSE + REQUIRED_USE 5.0.1 - Run on NVIDIA GPUs through tinygrad's CUDA device
local-ai sci-ml/bitsandbytes IUSE + REQUIRED_USE 0.50.2 - Build the CUDA quantization kernels (COMPUTE_BACKEND=cuda)
local-ai sci-ml/torchaudio metadata only - - GPU acceleration via CUDA
local-ai sci-ml/torchcodec metadata only - - NVDEC video decoding and nvJPEG via CUDA
masterlay media-gfx/blender IUSE + REQUIRED_USE 5.2.0 - -
masterlay media-gfx/openvdb IUSE 13.0.0 - -
masterlay media-libs/oidn IUSE 2.5.0 - -
mim media-gfx/blender IUSE + REQUIRED_USE 4.2.16 - -
mim media-libs/oidn IUSE 2.3.0 - -
mim media-libs/opencv IUSE + REQUIRED_USE 4.10.0 - Enable NVIDIA Cuda computations support (Experimental!)
mim media-libs/openimageio IUSE 3.0.7.0 - -
mim sci-libs/dlib IUSE 19.24.4 - Enable support for CUDA for Deep Neural Networks (cuDNN) on GPU (experimental)
mim sys-apps/hwloc IUSE 2.11.1 - Enable CUDA device discovery using libcudart
oubliette acct-user/ollama IUSE 3 - -
pentoo dev-libs/pocl IUSE 7.2 - -
pf4public sci-misc/llama-cpp IUSE 0.8.0, 0.9.0, 1.0.0, 9999 - Enable the NVIDIA CUDA GPU backend
pf4public sci-misc/stable-diffusion-cpp IUSE 1106, 1111, 9999 - Enable the NVIDIA CUDA GPU backend
pf4public sci-misc/strata IUSE + REQUIRED_USE 0.2.18, 0.2.19, 9999 - Build the engine for NVIDIA GPUs (CUDA backend, sm_75 through sm_120: RTX 20/30/40/50)
pf4public sci-misc/whisper-cpp IUSE 1.9.5, 9999 - Enable the NVIDIA CUDA backend (GGML_CUDA)
pf4public sci-ml/ggml IUSE 0.28.0, 0.29.0, 1.0.0, 9999 - Enable the NVIDIA CUDA GPU backend
pg_overlay media-libs/opencv IUSE + REQUIRED_USE 4.13.0, 5.0.0 - Enable NVIDIA Cuda computations support (Experimental!)
pg_overlay sci-mathematics/mathematica IUSE 14.2.1 - Install with cuda support
pingwho-overlay acct-user/ollama IUSE 2 - -
pingwho-overlay sci-ml/ollama-bin IUSE 0.40.17, 9999 - Support for Nvidia CUDA compatible cards
pkalin net-analyzer/suricata IUSE 7.0.10, 7.0.5, 7.0.8, 8.0.1 - Enable NVIDIA Cuda computations support
pypi dev-python/NewsFrames IUSE 1.3.7 - -
pypi dev-python/NewsSentiment IUSE 1.2.28 - -
pypi dev-python/absl-extra IUSE 0.1.3 - -
pypi dev-python/ctransformers IUSE 0.2.27 - -
pypi dev-python/datadm IUSE 0.3.6 - -
pypi dev-python/deface IUSE 1.5.0 - -
pypi dev-python/elephant IUSE 1.1.1 - -
pypi dev-python/gpjax-nightly IUSE 0.5.8, 0.5.9_pre20230314 - -
pypi dev-python/jax IUSE 0.8.3, 0.9.0 - -
pypi dev-python/jaxutils IUSE 0.0.8, 0.8 - -
pypi dev-python/libceed IUSE 0.12.0 - -
pypi dev-python/onnxruntime-gpu IUSE 1.23.2 - -
pypi dev-python/psyneulink IUSE 0.18.0.0 - -
pypi dev-python/py4DSTEM IUSE 0.14.18 - -
pypi dev-python/spacy IUSE 3.8.11, 4.0.0_pre3 - -
pypi dev-python/spacy-nightly IUSE 3.0.0_rc5 - -
pypi dev-python/spacy-transformers IUSE 1.3.9 - -
pypi dev-python/spacy-wrap IUSE 1.4.5 - -
pypi dev-python/thinc IUSE 8.3.10, 9.1.1 - -
pypi-sci dev-python/cupy IUSE + REQUIRED_USE 12.2.0, 13.2.0, 13.3.0, 14.0.1 - Enable support for nVidia CUDA
pypi-sci sci-libs/lightgbm IUSE 4.5.0 - -
raw net-analyzer/suricata IUSE 4.1.5, 5.0.2 - -
riru sci-libs/vtk IUSE + REQUIRED_USE 9.6.2 - -
ros-overlay sci-libs/pcl IUSE 1.9.1 - -
ruizhi-overlay dev-python/comfy-kitchen IUSE + REQUIRED_USE 0.2.41 - Build NVIDIA CUDA kernels
ruizhi-overlay media-gfx/comfyui IUSE + REQUIRED_USE 0.46.0 - Use CUDA-enabled PyTorch, torchvision and torchaudio operators; CPU is the default when neither GPU flag is enabled.
ruizhi-overlay sci-misc/prism-llama-cpp IUSE 12128, 9999 - -
ruizhi-overlay sci-ml/strata IUSE + REQUIRED_USE 0.2.18 - Build the engine for NVIDIA GPUs with the CUDA toolkit
ruizhi-overlay sci-ml/torchaudio IUSE 2.11.1 - Build CUDA audio operators to match a CUDA-enabled PyTorch.
sage-on-gentoo sci-libs/cholmod IUSE 5.3.4 - Use nvidia cuda toolkit for speeding up computations
salfter app-accessibility/whisper-cpp IUSE 1.9.4 - -
science dev-libs/boost-compute IUSE 1.74.0 - build cuda examples and tests
science sci-biology/fsl IUSE 6.0.4, 6.0.5 - Add support for CUDA
science sci-biology/megahit IUSE 1.2.9 - Enable NVIDIA GTX680 (4G memory) and Tesla K40c (12G memory) with CUDA 5.5, 6.0 and 6.5
science sci-chemistry/relion IUSE 4.0.0, 9999 - Enable CUDA support
science sci-libs/arrayfire IUSE 3.8.0 - Build CUDA backend.
science sci-libs/lightgbm IUSE 4.5.0 - -
science sci-libs/torchaudio IUSE + REQUIRED_USE 2.4.1 - -
science sys-devel/DPC++ IUSE + REQUIRED_USE 2021.12, 2022.09, 2022.12, 7.2.0 - use the cuda backend
seden acct-user/ollama IUSE 3 - -
sft dev-cpp/thrust IUSE 1.17.2 - -
sinustrom sci-ml/ggml IUSE + REQUIRED_USE 0.21.0 - -
stuff app-accessibility/whisper-cpp IUSE 1.9.3, 1.9.4, 1.9.5, 9999 - Enable NVIDIA CUDA GPU backend
stuff dev-python/comfy-aimdo-bin IUSE 0.6.0 - -
stuff dev-python/comfy-kitchen-bin IUSE 0.2.41 - -
stuff dev-python/cupy IUSE + REQUIRED_USE 13.6.0, 14.1.1, 14.2.0 - Enable support for nVidia CUDA
stuff dev-python/hyperspy IUSE 2.3.0, 2.4.0, 2.5.0 - -
stuff dev-python/sasmodels IUSE 1.0.12, 1.1.0 - Enable CUDA GPU-accelerated model evaluation via <pkg>dev-python/pycuda</pkg>
stuff dev-python/tilelang IUSE + REQUIRED_USE 0.1.10, 0.1.12, 0.1.13, 0.1.14, 0.1.15 - -
stuff dev-python/triton IUSE 3.7.1, 3.8.0, 3.9.0 - -
stuff dev-python/vllm IUSE + REQUIRED_USE 0.32.0, 0.33.0, 0.34.0 - -
stuff dev-python/xgrammar IUSE 0.2.7, 0.2.8, 0.2.9 - -
stuff media-gfx/comfyui IUSE + REQUIRED_USE 0.45.6, 0.46.0, 0.46.1 - Build the torch backend (<pkg>sci-ml/caffe2</pkg>) with CUDA/NVIDIA acceleration; also selects the CUDA wheels of <pkg>dev-python/comfy-aimdo-bin</pkg> and <pkg>dev-python/comfy-kitchen-bin</pkg> (otherwise their non-CUDA wheels)
stuff media-libs/opencv IUSE + REQUIRED_USE 4.14.0, 5.0.0, 5.1.0 - Enable NVIDIA Cuda computations support (Experimental!)
stuff sci-libs/cholmod IUSE 5.3.5 - Use nvidia cuda toolkit for speeding up computations
stuff sci-libs/onnxruntime IUSE 1.31.0, 1.31.1, 1.32.0 - Build the NVIDIA CUDA execution provider
stuff sci-libs/vtk IUSE + REQUIRED_USE 9.7.2 - -
stuff sci-misc/llama-cpp IUSE + REQUIRED_USE 0.8.0, 0.9.0, 1.0.0, 1.0.0_p12503, 1.0.0_p12532, 1.0.0_p12559, 9999 - -
stuff sci-misc/stable-diffusion-cpp IUSE 0_pre1091, 0_pre1104, 0_pre1106, 9999 - -
stuff sci-ml/bitsandbytes IUSE + REQUIRED_USE 0.50.0 - -
stuff sci-ml/caffe2 IUSE + REQUIRED_USE 2.13.0, 2.14.0, 2.14.1, 2.15.0 - -
stuff sci-ml/kineto IUSE 0.4.0_p20260925 - -
stuff sci-ml/ollama IUSE 0.40.13, 0.40.15, 0.40.16 - Build the CUDA (cuda_v13) llama-server GPU backend via <pkg>dev-util/nvidia-cuda-toolkit</pkg>
stuff sci-ml/pytorch IUSE + REQUIRED_USE 2.13.0, 2.14.0, 2.14.1, 2.15.0 - -
stuff sci-ml/sherpa-onnx IUSE 1.13.8, 2.0.0, 2.0.1, 2.0.2 - Build the NVIDIA CUDA backend (requires CUDA Toolkit and a compiler version supported by nvcc)
stuff sci-ml/torchcodec IUSE 0.15.0, 0.16.0, 0.17.0 - Build the NVIDIA CUDA decoding/encoding backend (libnpp + libnvrtc)
stuff sci-ml/torchvision IUSE + REQUIRED_USE 0.28.0, 0.29.0, 0.29.2 - -
tatsh-overlay sci-libs/vtk IUSE + REQUIRED_USE 9.7.2 - -
tatsh-overlay sci-ml/flash-attn IUSE 2.8.3_p2 - -
tatsh-overlay sci-ml/trellis-cpp IUSE 0.8.1 - -
touchfish-os sci-physics/lammps metadata only - - Enable cuda gpu computing support
ubuilds net-analyzer/suricata IUSE 7.0.13 - -
vklimovs sci-misc/llama-cpp IUSE 0.9.0, 9999 - -
vowstar media-libs/openimageio IUSE 3.0.6.1, 3.0.8.1 - -
vowstar sci-libs/vtk IUSE + REQUIRED_USE 9.6.2 - -
vowstar sci-mathematics/mathematica IUSE 14.3.0 - -
waebbl media-libs/opensubdiv IUSE 3.4.4 Enable NVIDIA CUDA Toolkit support through dev-util/nvidia-cuda-toolkit Enable NVIDIA CUDA Toolkit support through <pkg>dev-util/nvidia-cuda-toolkit</pkg>
zGentoo dev-python/comfy-aimdo IUSE 0.6.0 - -
zGentoo media-libs/opencv IUSE + REQUIRED_USE 4.14.0 - Enable NVIDIA Cuda computations support (Experimental!)
zGentoo sci-ml/ik_llama-cpp IUSE 0.0.4869 - -
zGentoo sci-ml/llama-cpp IUSE 12257 - -
zGentoo sci-ml/ollama IUSE 0.40.17 - -
zGentoo sci-ml/stable-diffusion-cpp IUSE 0.0.669, 0.0.841 - -
zGentoo virtual/caffe2 IUSE + REQUIRED_USE 2.15.0 - -
zGentoo virtual/comfyui IUSE + REQUIRED_USE 0.46.0 - -
zGentoo virtual/pytorch IUSE + REQUIRED_USE 2.15.0 - -