Julian Kemmerer

Talking about some kind of csv pipeline description `input AND output` capability As well as better summary of latency,fmax from the tool

Good question @suarezvictor These types of functions we have been writing in PipelineC are currently 1) (and potentially 2) in the below image  I say this because on twitter...

I see you ask about side effects - a quick sentence if you are interested: These functions can maintain state (making them style 1 only). But the state the maintain...

See https://github.com/JulianKemmerer/PipelineC/blob/master/ram.h for work around

Should be good for (FSM style arbitration, or not) sharing of N resources among M users.

https://google.github.io/xls/scheduling/#scheduling-to-minimize-pipeline-registers

Could this be done with AST transforms to IFs? Delay implementing synthesis-tool level case statement support/optimizations beyond simple compares and muxes?

Both @bartokon and @suarezvictor have said things along the lines of `do a binary search`. bartokon has some experimental plots of the #stages vs fmax curves we would be "searching-along"...

It is important to note `the search is not always get to the absolute fmax possible`. We should have a mode for that. But also more typical `get to this...

I dont think I understand `analyze pyrtl feedback with critical times and group wires to distinct speed categories and put registers there` I think that assumes way too much of...