* Upgrade presets for CUDA 13.2.1, cuDNN 9.21.1.3, NCCL 2.30.4, nvCOMP 5.2.0.10, TensorRT 10.16.1.11, Triton Inference Server 2.68.0

Author: saudet

.github/actions/deploy-ubuntu/action.yml

@@ -49,10 +49,10 @@ runs:
           export ARCH=arm64
           export ARCH_CUDA=sbsa
           export PREFIX=aarch64-linux-gnu
-          export CUDA=cuda-repo-rhel8-13-1-local-13.1.1_590.48.01-1.aarch64.rpm
-          export CUDNN=cuda-13-9.19.0.56-1.aarch64
-          export NCCL=2.29.3-1+cuda13.1.aarch64
-          export NVCOMP=nvcomp-linux-sbsa-5.1.0.21_cuda13-archive
+          export CUDA=cuda-repo-rhel8-13-2-local-13.2.1_595.58.03-1.aarch64.rpm
+          export CUDNN=cuda-13-9.21.1.3-1.aarch64
+          export NCCL=2.30.4-1+cuda13.2.aarch64
+          export NVCOMP=nvcomp-linux-sbsa-5.2.0.10_cuda13-archive
           export USERLAND_BUILDME="buildme --aarch64"
         elif [[ "$CI_DEPLOY_PLATFORM" == "linux-ppc64le" ]]; then
           export ARCH=ppc64el
@@ -71,10 +71,10 @@ runs:
           export ARCH=amd64
           export ARCH_CUDA=x86_64
           export PREFIX=x86_64-linux-gnu
-          export CUDA=cuda-repo-rhel8-13-1-local-13.1.1_590.48.01-1.x86_64.rpm
-          export CUDNN=cuda-13-9.19.0.56-1.x86_64
-          export NCCL=2.29.3-1+cuda13.1.x86_64
-          export NVCOMP=nvcomp-linux-x86_64-5.1.0.21_cuda13-archive
+          export CUDA=cuda-repo-rhel8-13-2-local-13.2.1_595.58.03-1.x86_64.rpm
+          export CUDNN=cuda-13-9.21.1.3-1.x86_64
+          export NCCL=2.30.4-1+cuda13.2.x86_64
+          export NVCOMP=nvcomp-linux-x86_64-5.2.0.10_cuda13-archive
         fi
         echo "ARCH=$ARCH" >> $GITHUB_ENV
         echo "PREFIX=$PREFIX" >> $GITHUB_ENV
@@ -176,7 +176,7 @@ runs:
 
         if [[ -n ${ARCH_CUDA:-} ]] && [[ -n ${CI_DEPLOY_NEED_CUDA:-} ]]; then
           echo Installing CUDA, cuDNN, nvCOMP, etc
-          curl -LO https://developer.download.nvidia.com/compute/cuda/13.1.1/local_installers/$CUDA
+          curl -LO https://developer.download.nvidia.com/compute/cuda/13.2.1/local_installers/$CUDA
           curl -LO https://developer.download.nvidia.com/compute/cuda/repos/rhel8/$ARCH_CUDA/libcudnn9-$CUDNN.rpm
           curl -LO https://developer.download.nvidia.com/compute/cuda/repos/rhel8/$ARCH_CUDA/libcudnn9-devel-$CUDNN.rpm
           curl -LO https://developer.download.nvidia.com/compute/cuda/repos/rhel8/$ARCH_CUDA/libcudnn9-headers-$CUDNN.rpm
@@ -185,9 +185,9 @@ runs:
 
           $SUDO rpm -i --force --ignorearch --nodeps $CUDA libcudnn*.rpm libnccl*.rpm
           rm -f *.rpm *.tgz *.txz *.tar.*
-          pushd /var/cuda-repo-rhel8-13-1-local/; $SUDO rpm -i --force --ignorearch --nodeps cuda*.rpm libc*.rpm libn*.rpm; $SUDO rm *.rpm; popd
-          $SUDO ln -snf /usr/local/cuda/targets/*/include /usr/local/cuda/include
-          $SUDO ln -snf /usr/local/cuda/targets/*/lib /usr/local/cuda/lib64
+          pushd /var/cuda-repo-rhel8-13-2-local/; $SUDO rpm -i --force --ignorearch --nodeps cuda*.rpm libc*.rpm libn*.rpm; $SUDO rm *.rpm; popd
+          $SUDO ln -snf $(ls -1d /usr/local/cuda/targets/*/include | tail -n 1) /usr/local/cuda/include
+          $SUDO ln -snf $(ls -1d /usr/local/cuda/targets/*/lib | tail -n 1) /usr/local/cuda/lib64
           $SUDO ln -sf /usr/local/cuda/lib64/ /usr/local/cuda/lib
           $SUDO ln -sf /usr/local/cuda/lib64/stubs/libcuda.so /usr/local/cuda/lib64/libcuda.so
           $SUDO ln -sf /usr/local/cuda/lib64/stubs/libnvidia-ml.so /usr/local/cuda/lib64/libnvidia-ml.so
@@ -228,8 +228,8 @@ runs:
         if [[ "$CI_DEPLOY_PLATFORM" == "linux-arm64" ]] && [[ "$CI_DEPLOY_MODULE" == "tensorrt" ]]; then
           echo Installing TensorRT
           # python3 -m gdown 1LZRCv4ZAGiDQAu4pvADJIGntq4cGl5tU
-          curl -LO https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.15.1/tars/TensorRT-10.15.1.29.Linux.aarch64-gnu.cuda-13.1.tar.gz
-          $SUDO tar -hxvf TensorRT-10.15.1.29.Linux.aarch64-gnu.cuda-13.1.tar.gz -C /usr/local/
+          curl -LO https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.16.1/tars/TensorRT-10.16.1.11.Linux.aarch64-gnu.cuda-13.2.tar.gz
+          $SUDO tar -hxvf TensorRT-10.16.1.11.Linux.aarch64-gnu.cuda-13.2.tar.gz -C /usr/local/
           $SUDO ln -sf /usr/local/TensorRT* /usr/local/tensorrt
           $SUDO patchelf --set-rpath '$ORIGIN' /usr/local/tensorrt/lib/libnvinfer.so
           # Remove downloaded archives and unused libraries to avoid running out of disk space
@@ -239,8 +239,8 @@ runs:
         if [[ "$CI_DEPLOY_PLATFORM" == "linux-x86_64" ]] && [[ "$CI_DEPLOY_MODULE" == "tensorrt" ]]; then
           echo Installing TensorRT
           # python3 -m gdown 1dVhD-DEYY42QbZe1GXl-vxe3k6KqWGsL
-          curl -LO https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.15.1/tars/TensorRT-10.15.1.29.Linux.x86_64-gnu.cuda-13.1.tar.gz
-          $SUDO tar -hxvf TensorRT-10.15.1.29.Linux.x86_64-gnu.cuda-13.1.tar.gz -C /usr/local/
+          curl -LO https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.16.1/tars/TensorRT-10.16.1.11.Linux.x86_64-gnu.cuda-13.2.tar.gz
+          $SUDO tar -hxvf TensorRT-10.16.1.11.Linux.x86_64-gnu.cuda-13.2.tar.gz -C /usr/local/
           $SUDO ln -sf /usr/local/TensorRT* /usr/local/tensorrt
           $SUDO patchelf --set-rpath '$ORIGIN' /usr/local/tensorrt/lib/libnvinfer.so
           # Remove downloaded archives and unused libraries to avoid running out of disk space

.github/actions/deploy-windows/action.yml

@@ -139,34 +139,34 @@ runs:
 
         if "%CI_DEPLOY_PLATFORM%"=="windows-x86_64" if not "%CI_DEPLOY_NEED_CUDA%"=="" (
           echo Installing CUDA, cuDNN, nvCOMP, etc
-          curl -LO https://developer.download.nvidia.com/compute/cuda/13.1.1/local_installers/cuda_13.1.1_windows.exe
-          curl -LO https://developer.download.nvidia.com/compute/cudnn/redist/cudnn/windows-x86_64/cudnn-windows-x86_64-9.19.0.56_cuda13-archive.zip
+          curl -LO https://developer.download.nvidia.com/compute/cuda/13.2.1/local_installers/cuda_13.2.1_windows.exe
+          curl -LO https://developer.download.nvidia.com/compute/cudnn/redist/cudnn/windows-x86_64/cudnn-windows-x86_64-9.21.1.3_cuda13-archive.zip
           curl -LO http://www.winimage.com/zLibDll/zlib123dllx64.zip
-          curl -LO https://developer.download.nvidia.com/compute/nvcomp/redist/nvcomp/windows-x86_64/nvcomp-windows-x86_64-5.1.0.21_cuda13-archive.zip
-          cuda_13.1.1_windows.exe -s
-          unzip cudnn-windows-x86_64-9.19.0.56_cuda13-archive.zip
+          curl -LO https://developer.download.nvidia.com/compute/nvcomp/redist/nvcomp/windows-x86_64/nvcomp-windows-x86_64-5.2.0.10_cuda13-archive.zip
+          cuda_13.2.1_windows.exe -s
+          unzip cudnn-windows-x86_64-9.21.1.3_cuda13-archive.zip
           unzip zlib123dllx64.zip
-          unzip nvcomp-windows-x86_64-5.1.0.21_cuda13-archive.zip
-          move cudnn-windows-x86_64-9.19.0.56_cuda13-archive\bin\x64\*.dll "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\bin\x64"
-          move cudnn-windows-x86_64-9.19.0.56_cuda13-archive\include\*.h "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\include"
-          move cudnn-windows-x86_64-9.19.0.56_cuda13-archive\lib\x64\*.lib "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\lib\x64"
-          move dll_x64\zlibwapi.dll "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\bin\x64"
-          move nvcomp-windows-x86_64-5.1.0.21_cuda13-archive\include\* "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\include"
-          move nvcomp-windows-x86_64-5.1.0.21_cuda13-archive\include\nvcomp "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\include"
-          move nvcomp-windows-x86_64-5.1.0.21_cuda13-archive\bin\nvcomp*.dll "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\bin\x64"
-          move nvcomp-windows-x86_64-5.1.0.21_cuda13-archive\lib\nvcomp*.lib "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\lib\x64"
+          unzip nvcomp-windows-x86_64-5.2.0.10_cuda13-archive.zip
+          move cudnn-windows-x86_64-9.21.1.3_cuda13-archive\bin\x64\*.dll "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\bin\x64"
+          move cudnn-windows-x86_64-9.21.1.3_cuda13-archive\include\*.h "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\include"
+          move cudnn-windows-x86_64-9.21.1.3_cuda13-archive\lib\x64\*.lib "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\lib\x64"
+          move dll_x64\zlibwapi.dll "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\bin\x64"
+          move nvcomp-windows-x86_64-5.2.0.10_cuda13-archive\include\* "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\include"
+          move nvcomp-windows-x86_64-5.2.0.10_cuda13-archive\include\nvcomp "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\include"
+          move nvcomp-windows-x86_64-5.2.0.10_cuda13-archive\bin\nvcomp*.dll "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\bin\x64"
+          move nvcomp-windows-x86_64-5.2.0.10_cuda13-archive\lib\nvcomp*.lib "%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\lib\x64"
 
           rem echo Applying hotfix to Visual Studio 2019 for CUDA
           rem curl -LO https://raw.githubusercontent.com/microsoft/STL/main/stl/inc/cmath
           rem bash -c "find 'C:/Program Files (x86)/Microsoft Visual Studio/2019/Enterprise/VC/' -name cmath -exec cp -v cmath {} \;"
-          bash -c "sed -i 's/cublas_v2.h/cublas_api.h/g' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/cublasXt.h' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/cusolverDn.h' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/cusolverSp.h'"
-          bash -c "sed -i 's/#if !defined(__cplusplus) || __cplusplus >= 201103L/#if 1/g' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/driver_types.h'"
-          rem bash -c "sed -i '/block_merge_sort.cuh/d' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/cub/cub.cuh'"
-          rem bash -c "sed -i '/device_merge_sort.cuh/d' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/cub/cub.cuh'"
-          rem bash -c "sed -i '/device_segmented_sort.cuh/d' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/cub/cub.cuh'"
-          rem bash -c "sed -i '/warp_merge_sort.cuh/d' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/cub/cub.cuh'"
-          bash -c "cp -a ""C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/cccl/*"" ""C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/include/"""
-          bash -c "find 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.1/'"
+          bash -c "sed -i 's/cublas_v2.h/cublas_api.h/g' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/cublasXt.h' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/cusolverDn.h' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/cusolverSp.h'"
+          bash -c "sed -i 's/#if !defined(__cplusplus) || __cplusplus >= 201103L/#if 1/g' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/driver_types.h'"
+          rem bash -c "sed -i '/block_merge_sort.cuh/d' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/cub/cub.cuh'"
+          rem bash -c "sed -i '/device_merge_sort.cuh/d' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/cub/cub.cuh'"
+          rem bash -c "sed -i '/device_segmented_sort.cuh/d' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/cub/cub.cuh'"
+          rem bash -c "sed -i '/warp_merge_sort.cuh/d' 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/cub/cub.cuh'"
+          bash -c "cp -a ""C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/cccl/*"" ""C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/include/"""
+          bash -c "find 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.2/'"
         )
 
         if "%CI_DEPLOY_MODULE%"=="nvcodec" (
@@ -179,9 +179,9 @@ runs:
         if "%CI_DEPLOY_MODULE%"=="tensorrt" (
           echo Installing TensorRT
           rem python -m gdown 1GfmJ1BKbacLpUU-0i_mGu0sjrAS0Xzzi
-          curl -LO https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.15.1/zip/TensorRT-10.15.1.29.Windows.amd64.cuda-13.1.zip
-          unzip TensorRT-10.15.1.29.Windows.amd64.cuda-13.1.zip
-          move TensorRT-10.15.1.29 "%ProgramFiles%\NVIDIA GPU Computing Toolkit\TensorRT"
+          curl -LO https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.16.1/zip/TensorRT-10.16.1.11.Windows.amd64.cuda-13.2.zip
+          unzip TensorRT-10.16.1.11.Windows.amd64.cuda-13.2.zip
+          move TensorRT-10.16.1.11 "%ProgramFiles%\NVIDIA GPU Computing Toolkit\TensorRT"
         )
 
         if "%CI_DEPLOY_MODULE%"=="mkl" (
@@ -260,10 +260,10 @@ runs:
           C:/msys64/usr/bin/bazel.exe version
         )
         if exist "%ProgramFiles%\NVIDIA GPU Computing Toolkit" (
-          set "CUDA_PATH=%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1"
-          set "CUDA_PATH_V13_1=%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1"
-          set "PATH=%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\bin;%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\libnvvp;%PATH%"
-          echo CUDA Version 13.1.1>"%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.1\version.txt"
+          set "CUDA_PATH=%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2"
+          set "CUDA_PATH_V13_2=%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2"
+          set "PATH=%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\bin;%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\libnvvp;%PATH%"
+          echo CUDA Version 13.2.1>"%ProgramFiles%\NVIDIA GPU Computing Toolkit\CUDA\v13.2\version.txt"
         )
 
         set "CCACHE_DIR=%USERPROFILE%\ccache"

.github/workflows/cuda.yml

@@ -19,7 +19,7 @@ env:
   STAGING_REPOSITORY: ${{ secrets.STAGING_REPOSITORY }}
 jobs:
   linux-arm64:
-    runs-on: ubuntu-22.04
+    runs-on: ubuntu-22.04-arm
     strategy:
       matrix:
         file: [pom.xml, redist, redist-cublas, redist-cudnn, redist-cusolver, redist-cusparse, redist-npp, redist-nccl, redist-nvcomp]

.github/workflows/nvcodec.yml

@@ -19,7 +19,7 @@ env:
   STAGING_REPOSITORY: ${{ secrets.STAGING_REPOSITORY }}
 jobs:
   linux-arm64:
-    runs-on: ubuntu-22.04
+    runs-on: ubuntu-22.04-arm
     steps:
       - uses: bytedeco/javacpp-presets/.github/actions/deploy-ubuntu@actions
 #  linux-ppc64le:

.github/workflows/tensorrt.yml

@@ -19,7 +19,7 @@ env:
   STAGING_REPOSITORY: ${{ secrets.STAGING_REPOSITORY }}
 jobs:
   linux-arm64:
-    runs-on: ubuntu-22.04
+    runs-on: ubuntu-22.04-arm
     steps:
       - uses: bytedeco/javacpp-presets/.github/actions/deploy-ubuntu@actions
   linux-x86_64:

.github/workflows/tritonserver.yml

@@ -19,7 +19,7 @@ env:
 jobs:
   linux-x86_64:
     runs-on: ubuntu-22.04
-    container: nvcr.io/nvidia/tritonserver:26.01-pyt-python-py3
+    container: nvcr.io/nvidia/tritonserver:26.04-pyt-python-py3
     steps:
       - uses: bytedeco/javacpp-presets/.github/actions/deploy-ubuntu@actions
   platform:

CHANGELOG.md

@@ -4,7 +4,7 @@
  * Add new `SampleOnnxMNIST` in samples for TensorRT ([pull #1742](https://github.com/bytedeco/javacpp-presets/pull/1742))
  * Fix loading issues with `libomp.dylib` and `libiomp5.dylib` for DNNL and PyTorch on Mac
  * Include `model_package_loader.h` header file in presets for PyTorch ([issue #1729](https://github.com/bytedeco/javacpp-presets/issues/1729))
- * Upgrade presets for FFmpeg 8.1, OpenBLAS 0.3.32, CPython 3.14.4, NumPy 2.4.4, SciPy 1.17.1, LLVM 22.1.1, PyTorch 2.11.0, TensorFlow Lite 2.21.0, ONNX 1.21.0, ONNX Runtime 1.24.4 ([pull #1750](https://github.com/bytedeco/javacpp-presets/pull/1750)), and their dependencies
+ * Upgrade presets for FFmpeg 8.1, OpenBLAS 0.3.32, CUDA 13.2.1, cuDNN 9.21.1.3, NCCL 2.30.4, nvCOMP 5.2.0.10, CPython 3.14.4, NumPy 2.4.4, SciPy 1.17.1, LLVM 22.1.1, PyTorch 2.11.0, TensorFlow Lite 2.21.0, TensorRT 10.16.1.11, Triton Inference Server 2.68.0, ONNX 1.21.0, ONNX Runtime 1.24.4 ([pull #1750](https://github.com/bytedeco/javacpp-presets/pull/1750)), and their dependencies
  * Compile classes with `parameters` bumping minimum requirements to Java SE 8 and Android 7.0 ([issue #1739](https://github.com/bytedeco/javacpp-presets/issues/1739))
 
 ### February 22, 2026 version 1.5.13

README.md

@@ -217,19 +217,19 @@ Each child module in turn relies by default on the included [`cppbuild.sh` scrip
  * Tesseract 5.5.x  https://github.com/tesseract-ocr/tesseract
  * Caffe 1.0  https://github.com/BVLC/caffe
  * OpenPose 1.7.0  https://github.com/CMU-Perceptual-Computing-Lab/openpose
- * CUDA 13.1.x  https://developer.nvidia.com/cuda-downloads
-   * cuDNN 9.19.x  https://developer.nvidia.com/cudnn
-   * NCCL 2.29.x  https://developer.nvidia.com/nccl
-   * nvCOMP 5.1.x https://developer.nvidia.com/nvcomp
+ * CUDA 13.2.x  https://developer.nvidia.com/cuda-downloads
+   * cuDNN 9.21.x  https://developer.nvidia.com/cudnn
+   * NCCL 2.30.x  https://developer.nvidia.com/nccl
+   * nvCOMP 5.2.x https://developer.nvidia.com/nvcomp
  * NVIDIA Video Codec SDK 13.0.x  https://developer.nvidia.com/nvidia-video-codec-sdk
  * OpenCL 3.0.x  https://github.com/KhronosGroup/OpenCL-ICD-Loader
  * MXNet 1.9.x  https://github.com/apache/incubator-mxnet
  * PyTorch 2.11.x  https://github.com/pytorch/pytorch
  * SentencePiece 0.2.x  https://github.com/google/sentencepiece
  * TensorFlow 1.15.x  https://github.com/tensorflow/tensorflow
  * TensorFlow Lite 2.21.x  https://github.com/tensorflow/tensorflow
- * TensorRT 10.15.x  https://developer.nvidia.com/tensorrt
- * Triton Inference Server 2.64.x  https://developer.nvidia.com/nvidia-triton-inference-server
+ * TensorRT 10.16.x  https://developer.nvidia.com/tensorrt
+ * Triton Inference Server 2.68.x  https://developer.nvidia.com/nvidia-triton-inference-server
  * The Arcade Learning Environment 0.8.x  https://github.com/mgbellemare/Arcade-Learning-Environment
  * DepthAI 2.24.x  https://github.com/luxonis/depthai-core
  * ONNX 1.20.x  https://github.com/onnx/onnx