Using OpenCv with CUDA GPU and iGPU

Accelerated Computing GPU-Accelerated Libraries
, , ,
1

Hello everyone, I have been working for several years now with OPENCV and CUDA. (currently versions 4.8.0 and 12.2 respectively). Everything works fine in my different jobs but I’ve always been stuck on an element that I can’t find (I’m on Windows 11).

I have several monitors connected to the GPU which allows me to display my work. And I also have a monitor connected to the CPU.

GPU : NVIDIA GeForce RTX 3060 Ti
CPU : Intel(R) UHD Graphics 630

If I set through Windows that the monitor connected to the GPU has priority, everything works fine.
Whereas if I define my monitor connected to the CPU as a priority (under windows), I can no longer use my program.

cudaGraphicsGLRegisterBuffer return error 304

Here is my opencv configuration as needed.

thank you for your advice.

Zanti Jonathan

Device Name : .\DISPLAY1
Device String : NVIDIA GeForce RTX 3060 Ti
Device Location x : -5120
Device Location y : 0
Device Size largeur : 2560
Device Size hauteur : 1440
Device PRIMARY : NO
Device Name : .\DISPLAY2
Device String : NVIDIA GeForce RTX 3060 Ti
Device Location x : -2560
Device Location y : 0
Device Size largeur : 2560
Device Size hauteur : 1440
Device PRIMARY : NO
Device Name : .\DISPLAY5
Device String : Intel(R) UHD Graphics 630
Device Location x : 0
Device Location y : 0
Device Size largeur : 2560
Device Size hauteur : 1440
Device PRIMARY : YES

cuda::getCudaEnabledDeviceCount() → CUDA : 1
Device ID : 0
Device Name : NVIDIA GeForce RTX 3060 Ti

General configuration for OpenCV 4.8.0-dev =====================================

Version control: unknown
Extra modules:
Location (extra): G:/Creation OPEN CV 4.8.0-dev/opencv_contrib-4.x/modules
Version control (extra): unknown
Platform:
Timestamp: 2023-08-12T13:54:15Z
Host: Windows 10.0.19045 AMD64
CMake: 3.27.2
CMake generator: Visual Studio 17 2022
CMake build tool: C:/Program Files/Microsoft Visual studio/2022/Community/MSBuild/Current/Bin/amd64/MSBuild.exe
MSVC: 1936
Configuration: Release
CPU/HW features:
Baseline: SSE SSE2 SSE3
requested: SSE3
Dispatched code generation: SSE4_1 SSE4_2 FP16 AVX AVX2 AVX512_SKX
requested: SSE4_1 SSE4_2 AVX FP16 AVX2 AVX512_SKX
SSE4_1 (16 files): + SSSE3 SSE4_1
SSE4_2 (1 files): + SSSE3 SSE4_1 POPCNT SSE4_2
FP16 (0 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 AVX
AVX (3 files): + SSSE3 SSE4_1 POPCNT SSE4_2 AVX
AVX2 (30 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2
AVX512_SKX (3 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2 AVX_512F AVX512_COMMON AVX512_SKX
C/C++:
Built as dynamic libs?: YES
C++ standard: 11
C++ Compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.36.32532/bin/Hostx64/x64/cl.exe (ver 19.36.32534.0)
C++ flags (Release): /DWIN32 /D_WINDOWS /W4 /GR /D _CRT_SECURE_NO_DEPRECATE /D _CRT_NONSTDC_NO_DEPRECATE /D _SCL_SECURE_NO_WARNINGS /Gy /bigobj /Oi /fp:precise /EHa /wd4127 /wd4251 /wd4324 /wd4275 /wd4512 /wd4589 /wd4819 /MP /MD /O2 /Ob2 /DNDEBUG /Zi
C++ flags (Debug): /DWIN32 /D_WINDOWS /W4 /GR /D _CRT_SECURE_NO_DEPRECATE /D _CRT_NONSTDC_NO_DEPRECATE /D _SCL_SECURE_NO_WARNINGS /Gy /bigobj /Oi /fp:precise /EHa /wd4127 /wd4251 /wd4324 /wd4275 /wd4512 /wd4589 /wd4819 /MP /MDd /Zi /Ob0 /Od /RTC1
C Compiler: C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.36.32532/bin/Hostx64/x64/cl.exe*
C flags (Release): /DWIN32 /D_WINDOWS /W3 /D _CRT_SECURE_NO_DEPRECATE /D _CRT_NONSTDC_NO_DEPRECATE /D _SCL_SECURE_NO_WARNINGS /Gy /bigobj /Oi /fp:precise /MP /MD /O2 /Ob2 /DNDEBUG /Zi
C flags (Debug): /DWIN32 /D_WINDOWS /W3 /D _CRT_SECURE_NO_DEPRECATE /D _CRT_NONSTDC_NO_DEPRECATE /D _SCL_SECURE_NO_WARNINGS /Gy /bigobj /Oi /fp:precise /MP /MDd /Zi /Ob0 /Od /RTC1
Linker flags (Release): /machine:x64 /INCREMENTAL:NO /debug
Linker flags (Debug): /machine:x64 /debug /INCREMENTAL*
ccache: NO
Precompiled headers: NO
Extra dependencies: cudart_static.lib nppc.lib nppial.lib nppicc.lib nppidei.lib nppif.lib nppig.lib nppim.lib nppist.lib nppisu.lib nppitc.lib npps.lib cublas.lib cufft.lib -LIBPATH:C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.2/lib/x64
3rdparty dependencies:
OpenCV modules:
To be built: aruco bgsegm bioinspired calib3d ccalib core cudaarithm cudabgsegm cudacodec cudafeatures2d cudafilters cudaimgproc cudalegacy cudaobjdetect cudaoptflow cudastereo cudawarping cudev datasets dpm face features2d flann fuzzy gapi hfs highgui img_hash imgcodecs imgproc intensity_transform line_descriptor ml objdetect optflow phase_unwrapping photo plot quality rapid reg rgbd saliency shape stereo stitching structured_light superres surface_matching tracking ts video videoio videostab world xfeatures2d ximgproc xobjdetect xphoto
Disabled: dnn dnn_objdetect dnn_superres java_bindings_generator python_bindings_generator python_tests
Disabled by dependency: mcc text wechat_qrcode
Unavailable: alphamat cvv freetype hdf java julia matlab ovis python2 python3 sfm viz
Applications: apps
Documentation: NO
Non-free algorithms: YES
Windows RT support: NO
GUI:
Win32 UI: YES
OpenGL support: YES (opengl32 glu32)
Media I/O:
ZLib: build (ver 1.2.13)
JPEG: build-libjpeg-turbo (ver 2.1.3-62)
SIMD Support Request: NO
WEBP: build (ver encoder: 0x020f)
PNG: build (ver 1.6.37)
TIFF: build (ver 42 - 4.2.0)
JPEG 2000: build (ver 2.5.0)
OpenEXR: build (ver 2.3.0)
HDR: YES
SUNRASTER: YES
PXM: YES
PFM: YES
Video I/O:
FFMPEG: YES (find_package)
avcodec: YES (60.3.100)
avformat: YES (60.3.100)
avutil: YES (58.2.100)
swscale: YES (7.1.100)
avresample: NO
avdevice: YES (60.1.100)
DirectShow: YES
Media Foundation: YES
DXVA: YES
Intel Media SDK: YES (C:/Program Files (x86)/IntelSWTools/Intel(R) Media SDK 2021 R1/Software Development Kit/lib/x64/libmfx_vs2015.lib)
Parallel framework: Concurrency
Trace: YES (with Intel ITT)
Other third-party libraries:
Intel IPP: 2021.8 [2021.8.0]
at: G:/Creation OPEN CV 4.8.0-dev/opencv 4.8.0-dev/build/3rdparty/ippicv/ippicv_win/icv
Intel IPP IW: sources (2021.8.0)
at: G:/Creation OPEN CV 4.8.0-dev/opencv 4.8.0-dev/build/3rdparty/ippicv/ippicv_win/iw
Custom HAL: NO
Protobuf: build (3.19.1)
Flatbuffers: builtin/3rdparty (23.5.9)
NVIDIA CUDA: YES (ver 12.2, CUFFT CUBLAS FAST_MATH)
NVIDIA GPU arch: 60 61 70 75 80 86 89 90
NVIDIA PTX archs: 86
OpenCL: YES (NVD3D11)
Include path: G:/Creation OPEN CV 4.8.0-dev/opencv 4.8.0-dev/3rdparty/include/opencl/1.2
Link libraries: Dynamic load
Python (for build): NO

2

This is expected behavior. This error:

indicates you are trying to use OpenGL and CUDA interop. That means you want to use OpenGL and also CUDA, working together. The only way that is possible is if the OpenGL context and the CUDA context are on a NVIDIA GPU, not on a Intel GPU. When you set the primary display to be the Intel GPU, (i.e. connect your monitor to it) then the OpenGL context gets created on the intel GPU. This is not a supported configuration for CUDA/OpenGL interop.

3

Good morning,

Thank you for your reply.

I totally understand… But what drives me to find a solution anyway is that:
If I run my program with the monitor in priority over the GPU. And that once the initialization phase has been executed, I modify the windows parameters by putting the monitor connected to the iGPU as a priority.
Everything works correctly.

Which makes me understand that there must be a solution to initialize everything differently…

4

Sure, make sure both CUDA context and OpenGL context are on the NVIDIA GPU. The CUDA context shouldn’t be an issue; if you are using e.g. the CUDA runtime API, you can’t/won’t create a CUDA context on the intel GPU.

But the OpenGL context can be created in either place. There are CUDA sample codes that show how to make sure the OpenGL context is in the right place on windows.