Is there any plan to support Dedicated WQs in the future?

[b3ng1998]

No description provided.

[skuma24]

Can you please give more details on your requirements? The dedicated WQs are harder to use (e.g., when multiple processes are started or forked and they all try to subscribe to a DSA WQ). Please let us know why SWQs dont work for you.

[b3ng1998]

• There will be only one process in our scenario. And according to this table, DWQ has better performance, which is preferred. That's why we want to use DWQ.

• I conducted some tests following the guidelines provided at https://github.com/intel/dsa-perf-micros/blob/main/doc/sample_command_lines.rst to confirm DWQ's superior performance in the asynchronous offload model. However, I'm uncertain about the performance difference in the synchronous offload model. Could you estimate the performance gap between the two WQs in the synchronous offload model? Thanks.