Is it possible to allocate the SMs to kernel or kernelet

Accelerated Computing CUDA CUDA Programming and Performance
#1

Hi All,

I am new to cuda programming. I am curious to know that is it possible to allocate SMs or compute percentage to kernel? I was searching answer to this query but not able to figure it out.
I am using NVIDIA Corporation GK210GL [Tesla K80] with cuda 9 toolkit installed.

Thanks in advance,
Shubham

#2

not possible

#3

But you can check SM ID in a kernel and exit if it’s not the one. You’ll have to do it in all kernels running on the GPU obviously.

#4

Thanks a lot dropbysh83… I will explore this suggestion!!