lightvector
the program quit without any error
I compile on my own computer with env RTX5060+cudnn9.13+cuda13.0 it compile perfect but cannot run
the env list:
path: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\bin\x64;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v13.0\bin;C:\Windows\system32;C:\Windows;C:\Windows\System32\Wbem;C:\Windows\System32\WindowsPowerShell\v1.0;C:\Windows\System32\OpenSSH;C:\ProgramData\miniconda3;C:\ProgramData\miniconda3\Scripts;C:\ProgramData\miniconda3\Library\bin;C:\Program Files\dotnet;C:\Program Files\NVIDIA Corporation\Nsight Compute 2025.2.0;C:\Program Files (x86)\NVIDIA Corporation\PhysX\Common;C:\Program Files\Git\cmd;C:\Program Files (x86)\Windows Kits\10\Windows Performance Toolkit;C:\Program Files\CMake\bin;C:\Users\Administrator\AppData\Local\Microsoft\WindowsApps;C:\Users\Administrator\AppData\Local\Programs\Microsoft VS Code\bin;C:\Program Files\NVIDIA\CUDNN\v9.13\bin\13.0;
I change the CMakeLists.txt: ...... if(USE_BACKEND STREQUAL "CUDA") message(STATUS "-DUSE_BACKEND=CUDA, using CUDA backend.")
enable_language(CUDA)
set(CUDA_STANDARD 11) set(NEURALNET_BACKEND_SOURCES neuralnet/cudabackend.cpp neuralnet/cudautils.cpp neuralnet/cudahelpers.cu )
https://en.wikipedia.org/wiki/CUDA#GPUs_supported
set(CMAKE_CUDA_ARCHITECTURES 90)
if(CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 12.8)
set(CMAKE_CUDA_ARCHITECTURES 50 52 53 60 61 62 70 72 75 80 86 87 90 120)
set(CMAKE_CUDA_FLAGS "-Wno-deprecated-gpu-targets")
elseif(CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 12.0)
set(CMAKE_CUDA_ARCHITECTURES 50 52 53 60 61 62 70 72 75 80 86 87 90)
set(CMAKE_CUDA_FLAGS "-Wno-deprecated-gpu-targets")
elseif(CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 11.8)
set(CMAKE_CUDA_ARCHITECTURES 35 37 50 52 53 60 61 62 70 72 75 80 86 87 90)
set(CMAKE_CUDA_FLAGS "-Wno-deprecated-gpu-targets")
elseif(CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 11.5)
set(CMAKE_CUDA_ARCHITECTURES 35 37 50 52 53 60 61 62 70 72 75 80 86 87)
set(CMAKE_CUDA_FLAGS "-Wno-deprecated-gpu-targets")
elseif(CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 11.1)
set(CMAKE_CUDA_ARCHITECTURES 35 37 50 52 53 60 61 62 70 72 75 80 86)
set(CMAKE_CUDA_FLAGS "-Wno-deprecated-gpu-targets")
elseif(CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 11.0)
set(CMAKE_CUDA_ARCHITECTURES 35 37 50 52 53 60 61 62 70 72 75 80)
set(CMAKE_CUDA_FLAGS "-Wno-deprecated-gpu-targets")
elseif (CMAKE_CUDA_COMPILER_VERSION VERSION_GREATER_EQUAL 10.2)
set(CMAKE_CUDA_ARCHITECTURES 30 35 37 50 52 53 60 61 62 70 72 75)
set(CMAKE_CUDA_FLAGS "-Wno-deprecated-gpu-targets")
else()
message(WARNING "CUDA 10.2 or greater is recommended, but attempting to build anyways")
set(CMAKE_CUDA_ARCHITECTURES 30 37 53 70)
endif()
cmake output: Selecting Windows SDK version 10.0.26100.0 to target Windows 10.0.22631. The C compiler identification is MSVC 19.44.35217.0 The CXX compiler identification is MSVC 19.44.35217.0 Detecting C compiler ABI info Detecting C compiler ABI info - done Check for working C compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped Detecting C compile features Detecting C compile features - done Detecting CXX compiler ABI info Detecting CXX compiler ABI info - done Check for working CXX compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe - skipped Detecting CXX compile features Detecting CXX compile features - done Building 'katago' executable for GTP engine and other tools. -DUSE_BACKEND=CUDA, using CUDA backend. The CUDA compiler identification is NVIDIA 13.0.88 with host compiler MSVC 19.44.35217.0 Detecting CUDA compiler ABI info Detecting CUDA compiler ABI info - done Check for working CUDA compiler: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v13.0/bin/nvcc.exe - skipped Detecting CUDA compile features Detecting CUDA compile features - done Including Git revision in the compiled executable, specify -DNO_GIT_REVISION=1 to disable CMake Warning at CMakeLists.txt:478 (message): WARNING: libzip library was NOT found. KataGo should still work for GTP/matches/analysis if everything else is good, but selfplay for writing training data will not be possible.
Setting up build for Windows. Setting up build for MSVC. Configuring done (16.0s) Generating done (0.0s)
run output: "compile\katago.exe" gtp -model "weights\b18c384nbt-optimisticv13-s5971M.bin.gz" -config "katago-v1.16.3-cuda12.8-cudnn9.8.0-windows-x64+bs50\analysis_example.cfg" 0<test_commands.txt KataGo v1.16.3 Using Chinese rules initially, unless GTP/GUI overrides this
D:\2023-06-15-windows64+katago>pause Press any key to continue . . .
