在Ubuntu上安装CUDA和cuDNN以及验证安装步骤
本教程详细介绍了如何在Ubuntu操作系统上安装CUDA（NVIDIA的并行计算平台）和cuDNN（深度神经网络库），以及如何验证安装是否成功。通过按照这些步骤操作，您将能够配置您的系统以利用GPU加速深度学习和其他计算密集型任务。此外，还包括如何设置环境变量和编译运行示例代码以验证CUDA和cuDNN的正常运行。
- 安装 CUDA
- 通过网络仓库安装CUDA（适用于Ubuntu）
- 配置环境变量
- 验证安装
- 安装 cuDNN
- 验证 cuDNN
安装 CUDA
在安装CUDA之前，我们需要进行一些预安装操作。首先，您需要安装当前正在运行的内核的头文件和开发包。打开终端并执行以下命令：
sudo apt-get install linux-headers-$(uname -r)
接下来，您需要删除过时的签名密钥：
sudo apt-key del 7fa2af80
通过网络仓库安装CUDA（适用于Ubuntu）
新的CUDA存储库的GPG公钥是3bf863cc。您可以通过cuda-keyring包或手动方法将其添加到系统中，不建议使用apt-key命令。执行以下步骤：
- 安装新的
cuda-keyring包。根据您的系统版本替换$distro/$arch：
wget https://developer.download.nvidia.com/compute/cuda/repos/$distro/$arch/cuda-keyring_1.1-1_all.deb
sudo dpkg -i cuda-keyring_1.1-1_all.deb
$distro/$arch 应该根据以下选项之一进行替换：
ubuntu1604/x86_64：适用于 Ubuntu 16.04 64位版本。ubuntu1804/cross-linux-sbsa：适用于 Ubuntu 18.04 交叉编译版本（SBSA 架构）。ubuntu1804/ppc64el：适用于 Ubuntu 18.04 64位 PowerPC 架构版本。
*ubuntu1804/sbsa：适用于 Ubuntu 18.04 SBSA 架构版本。ubuntu1804/x86_64：适用于 Ubuntu 18.04 64位版本。ubuntu2004/cross-linux-aarch64：适用于 Ubuntu 20.04 交叉编译版本（AArch64 架构）。ubuntu2004/arm64：适用于 Ubuntu 20.04 64位 ARM 架构版本。ubuntu2004/cross-linux-sbsa：适用于 Ubuntu 20.04 交叉编译版本（SBSA 架构）。ubuntu2004/sbsa：适用于 Ubuntu 20.04 SBSA 架构版本。ubuntu2004/x86_64：适用于 Ubuntu 20.04 64位版本。ubuntu2204/sbsa：适用于 Ubuntu 22.04 SBSA 架构版本。ubuntu2204/x86_64：适用于 Ubuntu 22.04 64位版本。
根据您的Ubuntu版本和架构选择适当的替代项来执行相应的安装步骤。
- 更新Apt仓库缓存：
sudo apt-get update
- 安装 CUDA SDK:
您可以使用以下命令获取可用的CUDA包列表：
cat /var/lib/apt/lists/*cuda*Packages | grep "Package:"
或查看下方列表：
| Meta Package | Purpose |
|---|---|
| cuda | Installs all CUDA Toolkit and Driver packages. Handles upgrading to the next version of the cuda package when it’s released. |
| cuda-12-2 | Installs all CUDA Toolkit and Driver packages. Remains at version 12.1 until an additional version of CUDA is installed. |
| cuda-toolkit-12-2 | Installs all CUDA Toolkit packages required to develop CUDA applications. Does not include the driver. |
| cuda-toolkit-12 | Installs all CUDA Toolkit packages required to develop applications. Will not upgrade beyond the 12.x series toolkits. Does not include the driver. |
| cuda-toolkit | Installs all CUDA Toolkit packages required to develop applications. Handles upgrading to the next 12.x version of CUDA when it’s released. Does not include the driver. |
| cuda-tools-12-2 | Installs all CUDA command line and visual tools. |
| cuda-runtime-12-2 | Installs all CUDA Toolkit packages required to run CUDA applications, as well as the Driver packages. |
| cuda-compiler-12-2 | Installs all CUDA compiler packages. |
| cuda-libraries-12-2 | Installs all runtime CUDA Library packages. |
| cuda-libraries-dev-12-2 | Installs all development CUDA Library packages. |
| cuda-drivers | Installs all Driver packages. Handles upgrading to the next version of the Driver packages when they’re released. |
选择你需要的包进行安装，这里选择 cuda-11.8
sudo apt-get install cuda-11-8
此安装包中包含显卡驱动，安装过程中，会让你输入密码，请记住该密码，后面重启电脑进入 Perform MOK managment 会使用到。
- 安装完成后，重新启动系统：
sudo reboot
配置 Perform MOK managment
选择 Enroll MOK （注册）-> 选择 Continue -> 选择 Enroll the key -> 选择 Yes -> 键入步骤3中输入的密码->选择 Reboot 重启电脑，完成英伟达显卡驱动安装。
配置环境变量
- 使用
vim编辑~/.bashrc文件。
sudo vim ~/.bashrc
- 在文件结尾添加以下内容：
export PATH=/usr/local/cuda-11.8/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-11.8/lib64\${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
${PATH:+:${PATH}}是一个用于设置环境变量的 Bash Shell 中的特殊语法。它的作用是在添加新路径到环境变量时，确保如果原始变量（在这种情况下是$PATH）已经包含一些路径，那么新路径会添加在原有路径的末尾，而且它们之间会用冒号 : 分隔。
具体来说，${PATH:+:${PATH}}的含义是：
如果$PATH已经定义（非空），那么它会在新路径之前加上一个冒号 :，然后再添加新路径。
如果$PATH未定义或为空，那么它只会添加新路径，不会加冒号。
这个语法的目的是确保在向$PATH添加新路径时，保持路径之间用冒号分隔，以确保环境变量的正确格式。这在很多环境变量的设置中都很有用，因为它避免了路径之间缺少分隔符而导致的错误。
LD_LIBRARY_PATH 是一个环境变量，用于指定动态链接器（dynamic linker）在运行可执行文件时搜索共享库文件（动态链接库或共享对象文件）的路径。在 Linux 和类Unix系统中，共享库文件包含在各种程序中，允许多个程序共享相同的库，从而减少内存占用并提高系统的效率。
- 刷新配置
在终端中运行以下命令，以使新的环境变量设置生效：
source ~/.bashrc
验证安装
首先，我们需要安装一些CUDA示例所需的第三方库。这些示例通常会在构建过程中检测所需的库，但如果未检测到，您需要手动安装它们。打开终端并执行以下命令：
sudo apt-get install g++ freeglut3-dev build-essential libx11-dev \libxmu-dev libxi-dev libglu1-mesa libglu1-mesa-dev libfreeimage-dev
完成第三方库依赖安装后，从 github 下载 https://github.com/nvidia/cuda-samples 源代码。
下载完成后，可以使用以下命令编译：
cd cuda-sample
sudo make
注意切换到你安装 cuda 版本的分支，这里是 v11.8。
可以完成整个编译，那么说明安装过程没有问题了。
在源代码目录执行 ./bin/x86_64/linux/release/deviceQuery 命令,结果如下所示：
cheungxiongwei@root:~/Source/cuda-samples$ ./bin/x86_64/linux/release/deviceQuery
./bin/x86_64/linux/release/deviceQuery Starting...CUDA Device Query (Runtime API) version (CUDART static linking)Detected 1 CUDA Capable device(s)Device 0: "NVIDIA GeForce RTX 4060 Laptop GPU"CUDA Driver Version / Runtime Version 12.2 / 11.8CUDA Capability Major/Minor version number: 8.9Total amount of global memory: 7940 MBytes (8325824512 bytes)
MapSMtoCores for SM 8.9 is undefined. Default to use 128 Cores/SM
MapSMtoCores for SM 8.9 is undefined. Default to use 128 Cores/SM(024) Multiprocessors, (128) CUDA Cores/MP: 3072 CUDA CoresGPU Max Clock rate: 2250 MHz (2.25 GHz)Memory Clock rate: 8001 MhzMemory Bus Width: 128-bitL2 Cache Size: 33554432 bytesMaximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layersMaximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layersTotal amount of constant memory: 65536 bytesTotal amount of shared memory per block: 49152 bytesTotal shared memory per multiprocessor: 102400 bytesTotal number of registers available per block: 65536Warp size: 32Maximum number of threads per multiprocessor: 1536Maximum number of threads per block: 1024Max dimension size of a thread block (x,y,z): (1024, 1024, 64)Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)Maximum memory pitch: 2147483647 bytesTexture alignment: 512 bytesConcurrent copy and kernel execution: Yes with 2 copy engine(s)Run time limit on kernels: YesIntegrated GPU sharing Host Memory: NoSupport host page-locked memory mapping: YesAlignment requirement for Surfaces: YesDevice has ECC support: DisabledDevice supports Unified Addressing (UVA): YesDevice supports Managed Memory: YesDevice supports Compute Preemption: YesSupports Cooperative Kernel Launch: YesSupports MultiDevice Co-op Kernel Launch: YesDevice PCI Domain ID / Bus ID / location ID: 0 / 1 / 0Compute Mode:< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.2, CUDA Runtime Version = 11.8, NumDevs = 1
Result = PASS
安装 cuDNN
安装 cuDNN库和 cuDNN 示例
sudo apt-get install libcudnn8=${cudnn_version}-1+${cuda_version}
sudo apt-get install libcudnn8-dev=${cudnn_version}-1+${cuda_version}
sudo apt-get install libcudnn8-samples=${cudnn_version}-1+${cuda_version}
根据以下内容进行替换:
${cudnn_version} is 8.9.4.*
${cuda_version} is cuda12.2 or cuda11.8
使用以下命令查找与 cuDNN 版本 “libcudnn8” 相关的软件包信息
cat /var/lib/apt/lists/*cuda*Packages | grep "./libcudnn8"
输出结果如下所示：
cheungxiongwei@root:~/cudnn_samples_v8/mnistCUDNN$ cat /var/lib/apt/lists/*cuda*Packages | grep "./libcudnn8"
Filename: ./libcudnn8_8.5.0.96-1+cuda11.7_amd64.deb
Filename: ./libcudnn8-dev_8.5.0.96-1+cuda11.7_amd64.deb
Filename: ./libcudnn8_8.6.0.163-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-dev_8.6.0.163-1+cuda11.8_amd64.deb
Filename: ./libcudnn8_8.7.0.84-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-dev_8.7.0.84-1+cuda11.8_amd64.deb
Filename: ./libcudnn8_8.8.0.121-1+cuda11.8_amd64.deb
Filename: ./libcudnn8_8.8.0.121-1+cuda12.0_amd64.deb
Filename: ./libcudnn8-dev_8.8.0.121-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-dev_8.8.0.121-1+cuda12.0_amd64.deb
Filename: ./libcudnn8_8.8.1.3-1+cuda11.8_amd64.deb
Filename: ./libcudnn8_8.8.1.3-1+cuda12.0_amd64.deb
Filename: ./libcudnn8-dev_8.8.1.3-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-dev_8.8.1.3-1+cuda12.0_amd64.deb
Filename: ./libcudnn8_8.9.0.131-1+cuda11.8_amd64.deb
Filename: ./libcudnn8_8.9.0.131-1+cuda12.1_amd64.deb
Filename: ./libcudnn8-dev_8.9.0.131-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-dev_8.9.0.131-1+cuda12.1_amd64.deb
Filename: ./libcudnn8_8.9.1.23-1+cuda11.8_amd64.deb
Filename: ./libcudnn8_8.9.1.23-1+cuda12.1_amd64.deb
Filename: ./libcudnn8-dev_8.9.1.23-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-dev_8.9.1.23-1+cuda12.1_amd64.deb
Filename: ./libcudnn8-samples_8.9.1.23-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-samples_8.9.1.23-1+cuda12.1_amd64.deb
Filename: ./libcudnn8_8.9.2.26-1+cuda11.8_amd64.deb
Filename: ./libcudnn8_8.9.2.26-1+cuda12.1_amd64.deb
Filename: ./libcudnn8-dev_8.9.2.26-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-dev_8.9.2.26-1+cuda12.1_amd64.deb
Filename: ./libcudnn8-samples_8.9.2.26-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-samples_8.9.2.26-1+cuda12.1_amd64.deb
Filename: ./libcudnn8_8.9.3.28-1+cuda11.8_amd64.deb
Filename: ./libcudnn8_8.9.3.28-1+cuda12.1_amd64.deb
Filename: ./libcudnn8-dev_8.9.3.28-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-dev_8.9.3.28-1+cuda12.1_amd64.deb
Filename: ./libcudnn8-samples_8.9.3.28-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-samples_8.9.3.28-1+cuda12.1_amd64.deb
Filename: ./libcudnn8_8.9.4.25-1+cuda11.8_amd64.deb
Filename: ./libcudnn8_8.9.4.25-1+cuda12.2_amd64.deb
Filename: ./libcudnn8-dev_8.9.4.25-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-dev_8.9.4.25-1+cuda12.2_amd64.deb
Filename: ./libcudnn8-samples_8.9.4.25-1+cuda11.8_amd64.deb
Filename: ./libcudnn8-samples_8.9.4.25-1+cuda12.2_amd64.deb
这里选择最新的 cudnn 8.9.4.25，和 cuda 11.8 进行替换，替换后的完整指令如下所示：
sudo apt-get install libcudnn8=8.9.4.25-1+cuda11.8
sudo apt-get install libcudnn8-dev=8.9.4.25-1+cuda11.8
sudo apt-get install libcudnn8-samples=8.9.4.25-1+cuda11.8
验证 cuDNN
要验证 cuDNN 是否已安装并正常运行，请编译 `/usr/src/cudnn_samples_v8`` 目录中的 mnistCUDNN 示例。
- 复制 cuDNN 示例到当前用户目录
cp -r /usr/src/cudnn_samples_v8/ $HOME
- 移动到 cuDNN 示例目录中
cd $HOME/cudnn_samples_v8/mnistCUDNN
- 编译 cuDNN mnisiCUDNN 示例
$make clean && make
如报错没有找到 FreeImage.h 文件，请执行 `sudo apt-get install libfreeimage-dev`` 指令安装该依赖。
- 运行 mnistCUDNN 示例
./mnistCUDNN
如果 cuDNN 在您的 Linux 系统上正确安装并编译&运行，您将看到类似以下内容的消息：
heungxiongwei@root:~/cudnn_samples_v8/mnistCUDNN$ ./mnistCUDNN
Executing: mnistCUDNN
cudnnGetVersion() : 8904 , CUDNN_VERSION from cudnn.h : 8904 (8.9.4)
Host compiler version : GCC 11.4.0There are 1 CUDA capable devices on your machine :
device 0 : sms 24 Capabilities 8.9, SmClock 2250.0 Mhz, MemSize (Mb) 7940, MemClock 8001.0 Mhz, Ecc=0, boardGroupID=0
Using device 0Testing single precision
Loading binary file data/conv1.bin
Loading binary file data/conv1.bias.bin
Loading binary file data/conv2.bin
Loading binary file data/conv2.bias.bin
Loading binary file data/ip1.bin
Loading binary file data/ip1.bias.bin
Loading binary file data/ip2.bin
Loading binary file data/ip2.bias.bin
Loading image data/one_28x28.pgm
Performing forward propagation ...
Testing cudnnGetConvolutionForwardAlgorithm_v7 ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: -1.000000 time requiring 178432 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: -1.000000 time requiring 184784 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: -1.000000 time requiring 2057744 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnFindConvolutionForwardAlgorithm ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.010240 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: 0.010240 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.018432 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 0.032992 time requiring 178432 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: 0.047104 time requiring 2057744 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 0.051200 time requiring 184784 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnGetConvolutionForwardAlgorithm_v7 ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: -1.000000 time requiring 128848 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: -1.000000 time requiring 128000 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: -1.000000 time requiring 4656640 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: -1.000000 time requiring 2450080 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: -1.000000 time requiring 1433120 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnFindConvolutionForwardAlgorithm ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 0.049152 time requiring 4656640 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.051200 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 0.058368 time requiring 2450080 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: 0.063648 time requiring 1433120 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.065536 time requiring 128000 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: 0.130112 time requiring 128848 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Resulting weights from Softmax:
0.0000000 0.9999399 0.0000000 0.0000000 0.0000561 0.0000000 0.0000012 0.0000017 0.0000010 0.0000000
Loading image data/three_28x28.pgm
Performing forward propagation ...
Testing cudnnGetConvolutionForwardAlgorithm_v7 ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: -1.000000 time requiring 178432 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: -1.000000 time requiring 184784 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: -1.000000 time requiring 2057744 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnFindConvolutionForwardAlgorithm ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.007328 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.010240 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: 0.011264 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: 0.024576 time requiring 2057744 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 0.025600 time requiring 184784 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 0.026624 time requiring 178432 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnGetConvolutionForwardAlgorithm_v7 ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: -1.000000 time requiring 128848 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: -1.000000 time requiring 128000 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: -1.000000 time requiring 4656640 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: -1.000000 time requiring 2450080 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: -1.000000 time requiring 1433120 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnFindConvolutionForwardAlgorithm ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 0.025376 time requiring 2450080 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: 0.030720 time requiring 128848 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 0.036864 time requiring 4656640 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.051200 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: 0.063488 time requiring 1433120 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.065536 time requiring 128000 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Resulting weights from Softmax:
0.0000000 0.0000000 0.0000000 0.9999288 0.0000000 0.0000711 0.0000000 0.0000000 0.0000000 0.0000000
Loading image data/five_28x28.pgm
Performing forward propagation ...
Resulting weights from Softmax:
0.0000000 0.0000008 0.0000000 0.0000002 0.0000000 0.9999820 0.0000154 0.0000000 0.0000012 0.0000006 Result of classification: 1 3 5Test passed!Testing half precision (math in single precision)
Loading binary file data/conv1.bin
Loading binary file data/conv1.bias.bin
Loading binary file data/conv2.bin
Loading binary file data/conv2.bias.bin
Loading binary file data/ip1.bin
Loading binary file data/ip1.bias.bin
Loading binary file data/ip2.bin
Loading binary file data/ip2.bias.bin
Loading image data/one_28x28.pgm
Performing forward propagation ...
Testing cudnnGetConvolutionForwardAlgorithm_v7 ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: -1.000000 time requiring 4608 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: -1.000000 time requiring 28800 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: -1.000000 time requiring 178432 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: -1.000000 time requiring 184784 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: -1.000000 time requiring 2057744 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnFindConvolutionForwardAlgorithm ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.011264 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.021504 time requiring 28800 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 0.022592 time requiring 184784 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 0.025600 time requiring 178432 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: 0.033792 time requiring 2057744 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: 0.074752 time requiring 4608 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnGetConvolutionForwardAlgorithm_v7 ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: -1.000000 time requiring 1536 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: -1.000000 time requiring 64000 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: -1.000000 time requiring 4656640 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: -1.000000 time requiring 2450080 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: -1.000000 time requiring 1433120 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnFindConvolutionForwardAlgorithm ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 0.031744 time requiring 2450080 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 0.040960 time requiring 4656640 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.051168 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: 0.060416 time requiring 1433120 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.064512 time requiring 64000 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: 0.069632 time requiring 1536 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Resulting weights from Softmax:
0.0000001 1.0000000 0.0000001 0.0000000 0.0000563 0.0000001 0.0000012 0.0000017 0.0000010 0.0000001
Loading image data/three_28x28.pgm
Performing forward propagation ...
Testing cudnnGetConvolutionForwardAlgorithm_v7 ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: -1.000000 time requiring 4608 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: -1.000000 time requiring 28800 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: -1.000000 time requiring 178432 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: -1.000000 time requiring 184784 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: -1.000000 time requiring 2057744 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnFindConvolutionForwardAlgorithm ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.009216 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.012288 time requiring 28800 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 0.021312 time requiring 184784 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: 0.023552 time requiring 4608 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 0.024352 time requiring 178432 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: 0.029696 time requiring 2057744 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnGetConvolutionForwardAlgorithm_v7 ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: -1.000000 time requiring 1536 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: -1.000000 time requiring 64000 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: -1.000000 time requiring 4656640 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: -1.000000 time requiring 2450080 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: -1.000000 time requiring 1433120 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Testing cudnnFindConvolutionForwardAlgorithm ...
^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 0.025600 time requiring 2450080 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 0.035840 time requiring 4656640 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.051200 time requiring 0 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 7: 0.060416 time requiring 1433120 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.064512 time requiring 64000 memory
^^^^ CUDNN_STATUS_SUCCESS for Algo 1: 0.065536 time requiring 1536 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 6: -1.000000 time requiring 0 memory
^^^^ CUDNN_STATUS_NOT_SUPPORTED for Algo 3: -1.000000 time requiring 0 memory
Resulting weights from Softmax:
0.0000000 0.0000000 0.0000000 1.0000000 0.0000000 0.0000714 0.0000000 0.0000000 0.0000000 0.0000000
Loading image data/five_28x28.pgm
Performing forward propagation ...
Resulting weights from Softmax:
0.0000000 0.0000008 0.0000000 0.0000002 0.0000000 1.0000000 0.0000154 0.0000000 0.0000012 0.0000006 Result of classification: 1 3 5Test passed!
