I have implemented a CUDA program.
Although the result is wrong, it works in debug mode anyway.
But after executing this program, I get the message like below.
(My code is too long so it’s difficult to show you my code here.)
Do you know the reason of it?
Is it because of wrong memory access or shortage of memory?
Any advice will be very helpful for me.
It’s hard to tell without looking at source code. Maybe you forgot to sync after kernel call? Maybe you have failing kernel?
Check result code of each CUDA function and find failing one, then post more details here.