Holoscan SDK

The Holoscan SDK CUDA 13 Python Wheel is part of NVIDIA Holoscan, the AI sensor processing platform that combines hardware systems for low-latency sensor and network connectivity, optimized libraries for data processing and AI, and core microservices to run streaming, imaging, and other applications, from embedded to edge to cloud. It can be used to build streaming AI pipelines for a variety of domains, including Medical Devices, High Performance Computing at the Edge, Industrial Inspection and more.

The Core distribution includes holoscan.testing for standalone operator tests, with public typing and a separately linked native testing library. Install pytest separately to run the examples; no testing-only wheel or GPU Python array package is required. Plain import holoscan does not eagerly import testing.

Getting Started

Visit the Holoscan User Guide to get started with the Holoscan SDK.

Prerequisites

• Prerequisites for each supported platform are documented in the user guide. Note that the python wheels have many optional dependencies which you may install manually based on your needs (see compatibility matrix at the bottom).
• The Holoscan SDK python wheels are tested on Ubuntu 24.04. They are generally expected to work on any Linux distribution with glibc 2.39 or above (see output of ldd --version) and CUDA Runtime 13.0 or above.
• Python: 3.12 to 3.14

Troubleshooting

ERROR: Could not find a version that satisfies the requirement holoscan-core-cu13==<version>
ERROR: No matching distribution found for holoscan-core-cu13==<version>

The latest version of the wheels was built and tested on Ubuntu 24.04 with glibc 2.39. You may need to switch to a Linux distribution with a more recent version of glibc (check your version with ldd --version), or use the Holoscan SDK NGC container instead.

libc.so.6: version 'GLIBC_2.32 not found
libstdc++.so.6: version `GLIBCXX_3.4.29` not found

Same as above.

ImportError: libcudart.so.13: cannot open shared object file: No such file or directory

CUDA runtime is missing from your system (required even for CPU only pipelines).

• x86_64: two options

  • A) System Installation: Follow the official installation steps for installing the CUDA Toolkit.
  • B) PIP installation:

    • For holoscan-core-cu13:

      python3 -m pip install nvidia-cuda-runtime==13.*
      

    • Export the CUDA runtime library path:

      export CUDA_WHL_LIB_DIR=$(python3 -c 'import nvidia.cuda_runtime; print(nvidia.cuda_runtime.__path__[0])')/lib
      export LD_LIBRARY_PATH="$CUDA_WHL_LIB_DIR:$LD_LIBRARY_PATH"
      

• Jetson Thor: Re-install JetPack 7.0.

• IGX Orin: holoscan-core-cu13 does not support IGX OS 1.x.

• Jetson Orin: holoscan-core-cu13 does not support Jetpack 6.

catastrophic error: cannot open source file "vector_types.h"

CUDA Runtime headers are missing from your system.

Resolution: same as above.

Reference: https://docs.cupy.dev/en/latest/install.html#cupy-always-raises-nvrtc-error-compilation-6

Error: libnvinfer.so.8: cannot open shared object file: No such file or directory
...
Error: libnvonnxparser.so.8: cannot open shared object file: No such file or directory

TensorRT is missing from your system (note that it is only needed by the holoscan.operators.InferenceOp operator.).

• x86_64:

  • A) System Installation: Follow the official installation steps.

  • B) PIP installation:

    python3 -m pip install tensorrt-libs~=8.6.1 --index-url https://pypi.nvidia.com
    export TRT_WHL_LIB_DIR=$(python3 -c 'import tensorrt_libs; print(tensorrt_libs.__path__[0])')
    export CUDNN_WHL_LIB_DIR=$(python3 -c 'import nvidia.cudnn; print(nvidia.cudnn.__path__[0])')/lib
    export CUBLAS_WHL_LIB_DIR=$(python3 -c 'import nvidia.cublas; print(nvidia.cublas.__path__[0])')/lib
    export LD_LIBRARY_PATH="$TRT_WHL_LIB_DIR:$CUDNN_WHL_LIB_DIR:$CUBLAS_WHL_LIB_DIR:$LD_LIBRARY_PATH"
    

• Jetson Thor: Re-install JetPack 7.0.

• IGX Orin: holoscan-core-cu13 does not support IGX OS 1.x.

• Jetson Orin: holoscan-core-cu13 does not support Jetpack 6.