How is the laptop GPU able to do the rendering and execute a cuda program at the same time

[zt9465]

Thanks again for the help! I have searched the question on both the forum and stackoverflow, and did find many related discussions, is possible to confirm the following understanding:?

All kernels will be serialized on the GPU no matter if they are launched by the same or diffrent CPU processes. So no context switching normally between multiple kernels. However, with MPS, it’s possible to make kernels run concurrently on the GPU, while the context switching details are not specified.

For future readers, so related discussions I found:

Does CPU process wait when calling CUDA code

CUDA kernel launches are asynchronous. This means that the CPU thread initiating the kernel launch makes a call into a library which starts the GPU processing. This library routine returns control to the CPU thread before the kernel has actually begun executing. The CPU thread can continue processing your code at that point (any code you have written after the point of the kernel launch), while the GPU kernel is executing. Two or more CPU processes can share a GPU in default compute mode, th…

Running more than one CUDA applications on one GPU

GPU sharing among different application with different CUDA context

Hello, Let’s say I have single Tesla K80 card and I have two different applications to be run on GPU. The two applications are designed to process a large chunk of data and expected to run for hours. Now the question is if they are create and launch the CUDA kernel with different context at same time then will they be executed in serialized way i.e. till the first application finishes other will wait? Or the CUDA driver can run concurrent kernels from different applications may be using time b…