What exactly do coder config InputPipeline and OutputPipeline do?

I would like to optimized vhdl coder generated code for speed.

The code is a simple matlab function that is a linear stringing of a set of matlab functions. I think placing pipeline stages between each of these calls would improve speed. However, if I set InputPipeline and OutputPipeline for the mast routine, will it place pipeline stages just at the beginnign and end of the master block or also for every subfunction call?

What are my alternative options beside enabling automated and distributed pipelining (which do seem to help a little)?

Is the only way to string separately compiled code blocks by and outside vhdl glue code or can I do this somehow within matlab?

Answer 1

[Kiran Kintali]

In MATLAB to HDL workflow InputPipeline and OutputPipeline options insert pipelines on the whole function.

You can also control finer placement of pipelines using coder.hdl.pipeline pragrma in the MATLAB code.

You can enable the DistributedPipelining to distribute these pipelines within the design. See the MATLAB to HDL optimization page for examples on the topic.