Category : gpu

I have a profile on VTune and it shows something running on the GPU (the line highlighted with the pale blue dot in the attached screenshot). How can I debug what in my codebase is running on there? To clarify: when that highlighted line is expanded, it’s the nvoglv64.dll process eating up all that time, ..

Read more

I have been trying to set up CUDA computing under Julia for my RTX 2070 GPU and, so far, I did not get any errors related to failed CUDA initialization when executing CUDA-parallelized code. However, the parallelized computations seem surprisingly slow, so I launched Pkg.test("CUDA") from Julia to get some more insight into why that ..

Read more

I have just installed the nvidia CUDA toolkit on my fresh Ubuntu 20.04 installation. Nvcc compiles CUDA programs, and they run without crashing. However, none of the results are correct. Here is the output of the test script (deviceQuery) that Nvidia provides: ./deviceQuery Starting… CUDA Device Query (Runtime API) version (CUDART static linking) Detected 1 ..

Read more

I have a piece of code that is tested on various Ubuntu 18 and Ubuntu 20 servers. It worked fine. But while deploying the same code on a new Laptop with GeForce GTX 1650 SUPER we are getting the following exception. Openedterminate called after throwing an instance of ‘cv::Exception’ what(): OpenCV(4.5.1-dev) /home/user/opencv_build/opencv_contrib/modules/cudaimgproc/src/canny.cpp:140: error: (-215:Assertion failed) ..

Read more

I have ported an algorithm, which exhibits good parallel efficiency through OpenMP on CPU, to the GPU through the OpenMP target directive (targeting nvptx). The performance however is lacking, as I only get a <2.0 speed up compared to single core performance. I have tried to minimize data movement and optimize for memory coalescing. I ..

Read more