Intel OpenCL SDK: clGetKernelWorkGroupInfo return value

Intel OpenCL SDK: clGetKernelWorkGroupInfo return value

Hi 

1. I have a global work size of 1024 by 1024.

2. I set the local work size to 16 by 16. 

3. My CPU opnecl device has a maximum work-group-size of 8192.

4. I call clEnqueueNDRangeKernel with the desired local-work-size (along with all other necessary parameters)

5. I call:

      a. clGetKernelWorkGroupInfo(kernel, device, CL_KERNEL_WORK_GROUP_SIZE, sizeof(size_t), (void*)&workGroupSizeUsed, NULL);

      b. clGetKernelWorkGroupInfo(kernel, device, CL_KERNEL_PREFERRED_WORK_GROUP_SIZE_MULTIPLE, sizeof(size_t), (void*)&workGroupSizeUsed, NULL);

6. Both calls return 8192. How is this possible?

My expectation is 16 - the value that I passed to it. 

Any help?

2 post / 0 nuovi
Ultimo contenuto
Per informazioni complete sulle ottimizzazioni del compilatore, consultare l'Avviso sull'ottimizzazione

Querying clGetKernelWorkGroupInfo with CL_KERNEL_WORK_GROUP_SIZE returns the maximum work group size supported for that kernel as determined by its resource utilization (e.g., private, local memory) or kernel attribute max_work_group_size.  On the other hand, CL_KERNEL_PREFERRED_WORK_GROUP_SIZE_MULTIPLE is a performance hint often used to map to a device's underlying SIMD architecture.  The two may return the same value, and CPUs seem like they would benefit from large work group sizes more than small work group sizes.

Accedere per lasciare un commento.