Rachit Nigam

We should handle #864 when we fix this

@bcarlet expressed interest in implementing this pass! Let's chat if you'd like to do this @bcarlet!

@calebmkim @paili0628 do you think this would a good pass to show off with the new `static` abstractions? If so, let's add it to the new static milestone

Got it! It seems to fall into the bucket of various passes that discover parallelism from your program so maybe it should be wrapped in the broader umbrella of a...

Subsumed by #1722

It's been roughly a year since we've thought of this problem. We've frankly not run into this much at all but worth asking @andrewb1999 if AMC might benefit from something...

After #1338, I've realized that there is no good way to eliminate this pass in general. Instead, I think we should change it so that it doesn't create unnecessary groups....

Once we wrap up #1725, we can start working on this. The new interface is flexible enough to allow for modules to be pipelined.

I don't know when or how this issue will be addressed. Cider does emit a warning on repeated overflow so I'm considering this "solved", at least in the dynamic case.

@calebmkim we briefly talked about implementing a weight-stationary systolic array. If we end up doing that for the camera-ready version of the paper, let's close this issue!