calebmkim

@sampsyo can also give his own answer, but here is a minimum example of where I think the language-level guarantee would affect our compaction abilities. Suppose we have the following...

Ah, I see what you're saying. I still think there are some tricky cases here, though. For example, suppose we have: ``` @promotable(20) seq { @promotable(10) A; @promotable(5) B; static...

Thanks for the response. Running the design with Verilator produces the same inconsistency in results. > I think it somewhat makes sense but the big question is how do you...

Sounds good-- I thin this shouldn't be too difficult if we are end up implementing a general `with` operator.

Ah I see. Maybe this is something that would be better to implement once the static group stuff is more developed.

We never fully reached a conclusion, but rn I'm leaning towards it not needing to be solved. I explain why I think that [here](https://github.com/cucapra/calyx/discussions/1523#discussioncomment-6136468). (although in the comment, you mention...

This discussion is subsumed by the overall FSM optimizations discussion described in #2020.

Yeah (we should test empirically on different benchmarks) but my intuition is that OHE should be profitable apart from loops/other high cycle counts.

Right now, there is a very basic hueristic you can use using the syntax: `-x cell-share:bounds=x,y,z`. `x` denotes the number of times you can share a given combinational cell (e.g.,...

> I would not allow for the QuertType::Port at all because that would tie us to a particular FSM implementation strategy. I also think we should have something like: `query:...