CUDA11基础镜像构建

制作一个CUDA环境，把依赖的相关库进行安装，Dockerfile如下：

FROM ubuntu:18.04 as base

ENV NVARCH x86_64

ENV NVIDIA_REQUIRE_CUDA "cuda>=11.4 brand=tesla,driver>=418,driver<419 brand=tesla,driver>=450,driver<451"
ENV NV_CUDA_CUDART_VERSION 11.4.43-1
ENV NV_CUDA_COMPAT_PACKAGE cuda-compat-11-4

RUN apt-get update && apt-get install -y --no-install-recommends \
    gnupg2 curl ca-certificates && \
    curl -fsSL https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/${NVARCH}/3bf863cc.pub | apt-key add - && \
    echo "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/${NVARCH} /" > /etc/apt/sources.list.d/cuda.list && \
    apt-get purge --autoremove -y curl \
    && rm -rf /var/lib/apt/lists/*

ENV CUDA_VERSION 11.4.0

# For libraries in the cuda-compat-* package: https://docs.nvidia.com/cuda/eula/index.html#attachment-a
RUN apt-get update && apt-get install -y --no-install-recommends \
    cuda-cudart-11-4=${NV_CUDA_CUDART_VERSION} \
    ${NV_CUDA_COMPAT_PACKAGE} \
    && ln -s cuda-11.4 /usr/local/cuda && \
    rm -rf /var/lib/apt/lists/*

# Required for nvidia-docker v1
RUN echo "/usr/local/nvidia/lib" >> /etc/ld.so.conf.d/nvidia.conf \
    && echo "/usr/local/nvidia/lib64" >> /etc/ld.so.conf.d/nvidia.conf

ENV PATH /usr/local/nvidia/bin:/usr/local/cuda/bin:${PATH}
ENV LD_LIBRARY_PATH /usr/local/nvidia/lib:/usr/local/nvidia/lib64

# COPY NGC-DL-CONTAINER-LICENSE /

# nvidia-container-runtime
ENV NVIDIA_VISIBLE_DEVICES all
ENV NVIDIA_DRIVER_CAPABILITIES compute,utility

ENV NV_CUDA_LIB_VERSION 11.4.0-1

ENV NV_NVTX_VERSION 11.4.43-1
ENV NV_LIBNPP_VERSION 11.4.0.33-1
ENV NV_LIBNPP_PACKAGE libnpp-11-4=${NV_LIBNPP_VERSION}
ENV NV_LIBCUSPARSE_VERSION 11.6.0.43-1

ENV NV_LIBCUBLAS_PACKAGE_NAME libcublas-11-4
ENV NV_LIBCUBLAS_VERSION 11.5.2.43-1
ENV NV_LIBCUBLAS_PACKAGE ${NV_LIBCUBLAS_PACKAGE_NAME}=${NV_LIBCUBLAS_VERSION}

ENV NV_LIBNCCL_PACKAGE_NAME libnccl2
ENV NV_LIBNCCL_PACKAGE_VERSION 2.11.4-1
ENV NCCL_VERSION 2.11.4-1
ENV NV_LIBNCCL_PACKAGE ${NV_LIBNCCL_PACKAGE_NAME}=${NV_LIBNCCL_PACKAGE_VERSION}+cuda11.4


RUN apt-get update && apt-get install -y --no-install-recommends \
    cuda-libraries-11-4=${NV_CUDA_LIB_VERSION} \
    ${NV_LIBNPP_PACKAGE} \
    cuda-nvtx-11-4=${NV_NVTX_VERSION} \
    libcusparse-11-4=${NV_LIBCUSPARSE_VERSION} \
    ${NV_LIBCUBLAS_PACKAGE} \
    ${NV_LIBNCCL_PACKAGE} \
    && rm -rf /var/lib/apt/lists/*


ENV NV_CUDA_LIB_VERSION "11.4.0-1"

ENV NV_CUDA_CUDART_DEV_VERSION 11.4.43-1
ENV NV_NVML_DEV_VERSION 11.4.43-1
ENV NV_LIBCUSPARSE_DEV_VERSION 11.6.0.43-1
ENV NV_LIBNPP_DEV_VERSION 11.4.0.33-1
ENV NV_LIBNPP_DEV_PACKAGE libnpp-dev-11-4=${NV_LIBNPP_DEV_VERSION}

ENV NV_LIBCUBLAS_DEV_VERSION 11.5.2.43-1
ENV NV_LIBCUBLAS_DEV_PACKAGE_NAME libcublas-dev-11-4
ENV NV_LIBCUBLAS_DEV_PACKAGE ${NV_LIBCUBLAS_DEV_PACKAGE_NAME}=${NV_LIBCUBLAS_DEV_VERSION}

ENV NV_NVPROF_VERSION 11.4.43-1
ENV NV_NVPROF_DEV_PACKAGE cuda-nvprof-11-4=${NV_NVPROF_VERSION}

ENV NV_LIBNCCL_DEV_PACKAGE_NAME libnccl-dev
ENV NV_LIBNCCL_DEV_PACKAGE_VERSION 2.11.4-1
ENV NCCL_VERSION 2.11.4-1
ENV NV_LIBNCCL_DEV_PACKAGE ${NV_LIBNCCL_DEV_PACKAGE_NAME}=${NV_LIBNCCL_DEV_PACKAGE_VERSION}+cuda11.4

LABEL maintainer "NVIDIA CORPORATION <cudatools@nvidia.com>"

RUN apt-get update && apt-get install -y --no-install-recommends \
    cuda-cudart-dev-11-4=${NV_CUDA_CUDART_DEV_VERSION} \
    cuda-command-line-tools-11-4=${NV_CUDA_LIB_VERSION} \
    cuda-minimal-build-11-4=${NV_CUDA_LIB_VERSION} \
    cuda-libraries-dev-11-4=${NV_CUDA_LIB_VERSION} \
    cuda-nvml-dev-11-4=${NV_NVML_DEV_VERSION} \
    ${NV_NVPROF_DEV_PACKAGE} \
    ${NV_LIBNPP_DEV_PACKAGE} \
    libcusparse-dev-11-4=${NV_LIBCUSPARSE_DEV_VERSION} \
    ${NV_LIBCUBLAS_DEV_PACKAGE} \
    ${NV_LIBNCCL_DEV_PACKAGE} \
    && rm -rf /var/lib/apt/lists/*

# Keep apt from auto upgrading the cublas and nccl packages. See https://gitlab.com/nvidia/container-images/cuda/-/issues/88
RUN apt-mark hold ${NV_LIBCUBLAS_DEV_PACKAGE_NAME} ${NV_LIBNCCL_DEV_PACKAGE_NAME}  ${NV_LIBCUBLAS_PACKAGE_NAME} ${NV_LIBNCCL_PACKAGE_NAME}

ENV LIBRARY_PATH /usr/local/cuda/lib64/stubs

CUDA相关库介绍

libcuda.so

The CUDA Driver API library for low-level CUDA programming.

libcudart.so

The CUDA Runtime API library for high-level CUDA programming, on top of the CUDA Driver API.

libcublas.so

The cuBLAS library is an implementation of BLAS (Basic Linear Algebra Subprograms) on top of the NVIDIA CUDA runtime. It allows the user to access the computational resources of NVIDIA Graphics Processing Unit (GPU), but does not auto-parallelize across multiple GPUs.

To use the cuBLAS library, the application must allocate the required matrices and vectors in the GPU memory space, fill them with data, call the sequence of desired cuBLAS functions, and then upload the results from the GPU memory space back to the host. The cuBLAS library also provides helper functions for writing and retrieving data from the GPU.

libcusparse.so

The cuSPARSE library contains a set of basic linear algebra subroutines used for handling sparse matrices. It is implemented on top of the NVIDIA CUDA runtime (which is part of the CUDA Toolkit) and is designed to be called from C and C++. The library routines can be classified into four categories:

• Level 1: operations between a vector in sparse format and a vector in dense format

• Level 2: operations between a matrix in sparse format and a vector in dense format

• Level 3: operations between a matrix in sparse format and a set of vectors in dense format (which can also usually be viewed as a dense tall matrix)

• Conversion: operations that allow conversion between different matrix formats

libcusolver.so

The cuSOLVER library contains LAPACK-like functions in dense and sparse linear algebra, including linear solver, least-square solver and eigenvalue solver.

libcufft.so, libcufftw.so

The NVIDIA CUDA Fast Fourier Transform (FFT) product consists of two separate libraries: cuFFT and cuFFTW. The cuFFT library is designed to provide high performance on NVIDIA GPUs. The cuFFTW library is provided as porting tool to enable users of FFTW to start using NVIDIA GPUs with a minimum amount of effort.

The FFT is a divide-and-conquer algorithm for efficiently computing discrete Fourier transforms of complex or real-valued data sets. It is one of the most important and widely used numerical algorithms in computational physics and general signal processing. The cuFFT library provides a simple interface for computing FFTs on an NVIDIA GPU, which allows users to quickly leverage the floating-point power and parallelism of the GPU in a highly optimized and tested FFT library.

libcurand.so

The cuRAND library provides facilities that focus on the simple and efficient generation of high-quality pseudorandom and quasirandom numbers. A pseudorandom sequence of numbers satisfies most of the statistical properties of a truly random sequence but is generated by a deterministic algorithm. A quasirandom sequence of n-dimensional points is generated by a deterministic algorithm designed to fill an n-dimensional space evenly.

libnppc.so, libnppi.so, libnpps.so

NVIDIA NPP is a library of functions for performing CUDA accelerated processing. The initial set of functionality in the library focuses on imaging and video processing and is widely applicable for developers in these areas. NPP will evolve over time to encompass more of the compute heavy tasks in a variety of problem domains. The NPP library is written to maximize flexibility, while maintaining high performance.

NPP can be used in one of two ways:

• A stand-alone library for adding GPU acceleration to an application with minimal effort. Using this route allows developers to add GPU acceleration to their applications in a matter of hours.

• A cooperative library for interoperating with a developer’s GPU code efficiently.

Either route allows developers to harness the massive compute resources of NVIDIA GPUs, while simultaneously reducing development times.

libnvvm.so

The NVVM library is used by NVCC to compile CUDA binary code to run on NVIDIA GPUs.

libdevice.so

The libdevice library is a collection of NVVM bitcode functions that implement common functions for NVIDIA GPU devices, including math primitives and bit-manipulation functions. These functions are optimized for particular GPU architectures, and are intended to be linked with an NVVM IR module during compilation to PTX.

libcuinj32.so, libcuinj64.so

The CUDA internal libraries for profiling. Used by nvprof and the Visual Profiler.

libnvToolsExt.so

The NVIDIA Tools Extension Library.

库说明链接地址