Open cuixing158 opened 6 years ago
You can compile with Visual Studio 2013/2015, CUDA 8.0 and cudnn 5.1.
Thank you!
hello,there。i meet the same problems just like yours, can u tell me how to fix this trouble? thanks
This version supports Visual Studio 2013, CUDA 8.0 and cudnn 5.1 only. If you want to install with cuda9.0, please use official version https://github.com/BVLC/caffe . You need to do every things from beginning.
I am using CUDA 9.0 now. What you need to do is to change the setting in the CommonSettings.props
file from
<CudaVersion>8.0</CudaVersion>
to
<CudaVersion>9.0</CudaVersion>
or other versions you installed. Then close visual studio and re-open it.
The old boost library is not compatible with CUDA >9.0. Please download the new thirdparty.rar
file.
ok,thanks。i know the problems,i‘m using Visual Studio 2015...so,obviously i meet this trouble。Emmm...what should i do if using vs2015?and by the way ,i'v complie caffe in ubuntu16.04+cuda9.0+cudnn
I see that you have updated the “thirdparty.rar” file to support cuda9.0, but in turn CUDA 8.0 is compatible? @happynear
firstly, thank you for @happynear. @cuixing158 @whu-lyh @pluskid @cuixing158, i have built sucessfully caffe-window with cuda10.0 and cudnn7.3.0.
Hello,can I ask a few questions about built caffe-window with VS2017 and cuda10.0 @wuqingshan2010
@wuqingshan2010 I meet some problem with cuda10, can you share your solution with it?
我配置文件写好路径后,出现如下错误,是cuda9.0不适合呢? C:\caffe\caffe-windows-ms\windows\thirdparty\Boost\boost/config/compiler/nvcc.hpp(22): fatal error C1017: 无效的整数常量表达式 1> base_data_layer.cu 1>C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V140\BuildCustomizations\CUDA 9.0.targets(707,9): error MSB3721: 命令“"C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\bin\nvcc.exe" -gencode=arch=compute_52,code=\"sm_52,compute_52\" -gencode=arch=compute_60,code=\"sm_60,compute_60\" --use-local-env --cl-version 2015 -ccbin "C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\bin\x86_amd64" -x cu -I"C:\caffe\caffe-windows-ms\windows\libcaffe\....\src\" -I"C:\caffe\caffe-windows-ms\windows\libcaffe\....\3rdparty\" -I"C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\include" -I"C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\include" --keep-dir C:\caffe\caffe-windows-ms\windows..\Build\Int\libcaffe\x64\Release -maxrregcount=0 --machine 64 --compile -cudart static -Xcompiler "/wd 4819" -Xcudafe "--diag_suppress=exception_spec_override_incompat --diag_suppress=useless_using_declaration --diag_suppress=field_without_dll_interface" -D_SCL_SECURE_NO_WARNINGS -DGFLAGS_DLL_DECL= -DNDEBUG -D_SCL_SECURE_NO_WARNINGS -DUSE_OPENCV -DUSE_LEVELDB -DUSE_LMDB -DMATLAB_MEX_FILE -DWITH_PYTHON_LAYER -DBOOST_PYTHON_STATIC_LIB -DUSE_NCCL -DUSE_CUDNN -D_UNICODE -DUNICODE -Xcompiler "/EHsc /W1 /nologo /Ox /FS /Zi /MD " -o C:\caffe\caffe-windows-ms\windows..\Build\Int\libcaffe\x64\Release\conv_layer.cu.obj "C:\caffe\caffe-windows-ms\src\caffe\layers\conv_layer.cu"”已退出,返回代码为 2。![err](https://user-images.githubusercontent.com/22857777/37562992-317ebcb4-2ab1-11e8-902f-2209a3c65016.jpg)