Multithreaded call from cpu to gpu

Accelerated Computing CUDA CUDA Programming and Performance
#1

Hi all,

I have implemented a small code in which cpu program is multithreaded. All cpu threads have access to the same gpu. After opening a certain number of cpu threads I get error from gpu(where gpu have parallel call from cpu). Is this erroneous to foroward multithreaded call to gpu where I have only one gpu card.

Please help.

#2

I know that one host thread can access only one device in it’s life-time. However, I’m not sure whether you can access the same device through multiple cpu threads…