can single work items kernels run in parallel on same device

Honored Contributor

8 years ago

--- Quote Start ---

but you could try using CL_MEM_COPY_HOST_PTR instead of CL_MEM_USE_HOST_PTR

--- Quote End ---

This did not help.

--- Quote Start ---

adding the 'restrict' flag to your global variables in the kernel

--- Quote End ---

This did not help either.

However, this got me tinkering though. I did try changing the CL_MEM_READ_WRITE to CL_MEM_WRITE_ONLY.

This did work!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

$ bin/host 100000 4

Reprogramming device [0] with handle 1

Task:2 complete (4.529 ms)

Task:3 complete (4.556 ms)

Task:0 complete (4.559 ms)

Task:1 complete (4.561 ms)

Time: 4.563 ms (1.141 ms / kernel)

Sum 0-100000.000000 (step 1.000000) = 5000050000.000000

Thanks SO MUCH to nicolacdnll and fand for giving some new suggestions that FINALLY lead to a solution!

Forum Discussion