NVIDIA
NVIDIA
CUDA Deep Learning
Container
NVIDIA
NVIDIA
CUDA Deep Learning

CUDA is a parallel computing platform and programming model that enhances computing performance using NVIDIA GPUs. CUDA Deep Learning integrates networking and GPU-accelerated libraries like cuDNN, cuTensor, NCCL, HPC-x, and the CUDA Toolkit.

LayerLabelCreated
9b07deb30551e8b8848a0315b185b4530d58b11038a3f970ac8545465b9b2ec2CONFIG
Entrypoint /opt/nvidia/nvidia_entrypoint.sh
11/14/2024 10:32 PM UTC
44a59ed340afb152c8e1ee310e9c969639b10078558b85fc95cef98bd507f224COPY
/opt/amazon/aws-ofi-nccl /opt/amazon/aws-ofi-nccl
11/14/2024 10:32 PM UTC
22db7d4cfcfe1cd2651acdb4e6779602a184c7630b4aab487c8c4d47c95e85eeENV
LIBRARY_PATH=/usr/local/cuda/lib64/stubs:
11/14/2024 10:31 PM UTC
2f2a55c74f44b6c7a0de6ace8885365729703c45869d04b451bbfb353367aab4RUN
GDRCOPY_VERSION=2.4.1 HPCX_VERSION=2.21 RDMACORE_VERSION=39.0 MOFED_VERSION=5.4-rdmacore39.0 OPENUCX_VERSION=1.18.0 OPENMPI_VERSION=4.1.7 EFA_VERSION=1.34.0 AWS_OFI_NCCL_VERSION=1.12.1 TARGETARCH=amd64 export DEVEL=1 BASE=0 &&
  /nvidia/build-scripts/installNCU.sh &&
  /nvidia/build-scripts/installCUDA.sh &&
  /nvidia/build-scripts/installLIBS.sh &&
  if [ ! -f /etc/ld.so.conf.d/nvidia-tegra.conf ]; then /nvidia/build-scripts/installNCCL.sh; fi &&
  /nvidia/build-scripts/installCUDNN.sh &&
  /nvidia/build-scripts/installCUTENSOR.sh &&
  /nvidia/build-scripts/installTRT.sh &&
  /nvidia/build-scripts/installNSYS.sh &&
  /nvidia/build-scripts/installCUSPARSELT.sh &&
  if [ -f "/tmp/cuda-${_CUDA_VERSION_MAJMIN}.patch" ]; then patch -p0 < /tmp/cuda-${_CUDA_VERSION_MAJMIN}.patch; fi &&
  rm -f /tmp/cuda-*.patch
11/14/2024 10:31 PM UTC
7e31a3b0749ee82e76e261535be12089b7869ffd51f7bbbe9ab0603c412992fcCOPY
cuda-*.patch /tmp
11/14/2024 10:28 PM UTC
ede82e089a56291f00c067d63c9ddff92553862ee20091200215229529a3eeafENV
OMPI_MCA_coll_hcoll_enable=0
11/14/2024 10:28 PM UTC
5b8a709d9deebf233f9ae9a1fb5a2b38f5bd344032fdb86aebb2e0ef3fa84103ENV
OPAL_PREFIX=/opt/hpcx/ompi PATH=/usr/local/mpi/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/ucx/bin:/opt/amazon/efa/bin
11/14/2024 10:28 PM UTC
ef1b6a9d83c526dd28ab3c47a45b44e91dc131ed132514a00518ffde9fc211f6RUN
GDRCOPY_VERSION=2.4.1 HPCX_VERSION=2.21 RDMACORE_VERSION=39.0 MOFED_VERSION=5.4-rdmacore39.0 OPENUCX_VERSION=1.18.0 OPENMPI_VERSION=4.1.7 EFA_VERSION=1.34.0 AWS_OFI_NCCL_VERSION=1.12.1 TARGETARCH=amd64 cd /nvidia &&
  ( export DEBIAN_FRONTEND=noninteractive &&
  apt-get update &&
  apt-get install -y --no-install-recommends libibverbs1 libibverbs-dev librdmacm1 librdmacm-dev libibumad3 libibumad-dev ibverbs-utils ibverbs-providers &&
  rm -rf /var/lib/apt/lists/* &&
  rm $(dpkg-query -L libibverbs-dev librdmacm-dev libibumad-dev | grep "\(\.so\|\.a\)$") ) &&
  ( cd opt/gdrcopy/ &&
  dpkg -i libgdrapi_*.deb ) &&
  ( cp -r opt/hpcx /opt/ &&
  cp etc/ld.so.conf.d/hpcx.conf /etc/ld.so.conf.d/ &&
  ln -sf /opt/hpcx/ompi /usr/local/mpi &&
  ln -sf /opt/hpcx/ucx /usr/local/ucx &&
  sed -i 's/^\(hwloc_base_binding_policy\) = core$/\1 = none/' /opt/hpcx/ompi/etc/openmpi-mca-params.conf &&
  sed -i 's/^\(btl = self\)$/#\1/' /opt/hpcx/ompi/etc/openmpi-mca-params.conf ) &&
  ( if [ ! -f /etc/ld.so.conf.d/nvidia-tegra.conf ]; then cd opt/amazon/efa/ &&
  dpkg -i libfabric*.deb &&
  rm /opt/amazon/efa/lib/libfabric.a &&
  echo "/opt/amazon/efa/lib" > /etc/ld.so.conf.d/efa.conf; fi ) &&
  ldconfig
11/14/2024 10:28 PM UTC
3c03e6081bca9e6d5a26b686b0c086ab3e8fc9332ffae4e9b6bb27bc326cf295ARG
TARGETARCH=amd64
11/14/2024 10:28 PM UTC
6cf83b6c0f50811a61c1b67a7292f4642d61a62a3861978289d5ade4cc337259ENV
GDRCOPY_VERSION=2.4.1 HPCX_VERSION=2.21 MOFED_VERSION=5.4-rdmacore39.0 OPENUCX_VERSION=1.18.0 OPENMPI_VERSION=4.1.7 RDMACORE_VERSION=39.0 EFA_VERSION=1.34.0 AWS_OFI_NCCL_VERSION=1.12.1
11/14/2024 10:28 PM UTC