Thomas B. Preußer

Results 8 comments of Thomas B. Preußer

The only eager follow-up work I am aware of is the one by the team of [Kenji Kise](https://www.arch.cs.titech.ac.jp/~kise/). They [published on further optimizing the solver design](https://ieeexplore.ieee.org/abstract/document/8540208) in 2018. However, rather...

Thanks for your feedback! Please, use the `unsigned` version of `resize` to make it semantically correct. Would you want to turn this fix into a pull request then?

`resize` on `std_logic_vector` is ill-defined as it has to make assumptions about signedness. I would rather fence in its use and put it on a deprecation path.

Yes, and it even has implied semantics using the index direction as a hint for alignment. I would now consider this another feature hard to convey. Although I admittedly tried,...

Mmhhh, the first of these uses would even be semantically correct ...

THanks, @vopade! This looks good. Getting ready for the merge ...

Thanks for this PR, @mdanilow! I need to check with the sister PR against FINN. This one forks off from a somewhat distant past. I'm not convinced yet that it...

This rectangular upsampling is superseded by the current `dev` version, which even supports arbitrary dimensionality beyond 2D rectangles. Feel free to re-open if there are any non-obvious functional limitations in...