Hey I’m trying to debug CUDA code on my system: I have a 970GTX (device 0) that I’m running the code on and a 750GTX (device 1) that I’m using for openGL and have the display plugged into. Code runs fine but just recently I’ve gotten some bugs that I need to debug on the GPU side. When I do the application just hangs even on sections of fully functioning code that was written, tested and debugged on my other system (which doesn’t support the features for the new code).

Is there some sort of setting in nsight that I have to set to tell it which GPU is doing the actual processing or is it something else?

Seems that its not the code that is running slow its nsight taking a long time to load “CUDA Modules” is there documentation where I can find out details about this process and why it seems to take so long?

I get this error when running memcheck for the command line:

========= CUDA-MEMCHECK
CUDA using device: GeForce GTX 970
pMesher Starting…

Reading file to buffer…
File Read: 12.024 sec
Starting GPU read opperation…
Building temp arrays

Line 520 CUDA Sync: unspecified launch failure

========= Internal Memcheck Error: Memcheck failed initialization as some other
tools is currently attached. Please make sure that nvprof and Nsight Visual Stud
io Edition are not being run simultaneously
========= Saved host backtrace up to driver entry point at error
========= Host Frame:C:\Windows\system32\nvcuda.dll (cuInit + 0x13d) [0x141d
]
========= Host Frame:C:\Users\aidan_000\Desktop\honours\pMesher\bin\Debug\pm
esher.exe (cudart::globalState::loadDriverInternal + 0x6c) [0x3242c]
========= Host Frame:C:\Users\aidan_000\Desktop\honours\pMesher\bin\Debug\pm
esher.exe (cudart::__loadDriverInternalUtil + 0x31) [0x34ef1]
========= Host Frame:C:\Users\aidan_000\Desktop\honours\pMesher\bin\Debug\pm
esher.exe (cudart::cuosOnce + 0x26) [0x40fc6]
========= Host Frame:C:\Users\aidan_000\Desktop\honours\pMesher\bin\Debug\pm
esher.exe (cudart::globalState::loadDriver + 0x2b) [0x34f2b]
========= Host Frame:C:\Users\aidan_000\Desktop\honours\pMesher\bin\Debug\pm
esher.exe (cudart::globalState::initializeDriver + 0x17) [0x35117]
========= Host Frame:C:\Users\aidan_000\Desktop\honours\pMesher\bin\Debug\pm
esher.exe (cudaGetDeviceCount + 0x36) [0x12256]
========= Host Frame:C:\Users\aidan_000\Desktop\honours\pMesher\bin\Debug\pm
esher.exe (main + 0x2c) [0x11f53c]
========= Host Frame:C:\Users\aidan_000\Desktop\honours\pMesher\bin\Debug\pm
esher.exe (__tmainCRTStartup + 0x19d) [0x12893d]
========= Host Frame:C:\Users\aidan_000\Desktop\honours\pMesher\bin\Debug\pm
esher.exe (mainCRTStartup + 0xe) [0x128a6e]
========= Host Frame:C:\Windows\system32\KERNEL32.DLL (BaseThreadInitThunk +
0x22) [0x13d2]
========= Host Frame:C:\Windows\SYSTEM32\ntdll.dll (RtlUserThreadStart + 0x3
4) [0x6e954]

========= ERROR SUMMARY: 1 error