Should we move to

GF180mcuD

? If so, there will be needed some modifications to the IO PADs? Klayout layers for GF180 are a little hard to visualize, I changed the

.lyp

file and had to modify a lot of layers to better visualize them, I hope that is OK.

I will get back to you on this very soon.

The issue is that I chose CG180mcuC pretty much at random from the selection of process variants that GF offered. Generally, the variant selection was based on reasonable assumptions about what we needed for caravel and the user projects. However, once we submitted GFMPW-1, GF told us that the top metal on that process variant is too thin and causes problems with delamination during wire bonding. So they advised us that they had created a different variant by increasing the top metal thickness from 0.9um to 1.1um, and that in the future we should use the 1.1um variant, which is what I called gf180mcuD in open_pdks. There is very little difference between the two variants and there are no differences in anything other than parasitics of the top level metal. I have a vague recollection of a via4 DRC rule change that I would have to look into because I don't remember if that was a real difference in the PDK or just an error that was in our klayout DRC deck. Generally speaking, though, you can take any design for gf180mcuC and rename it to gf180mcuD and you won't get any errors or find any appreciable difference.

Ok, I'm going to move to GF180mcuD, and also I'm going to re-check the DRC and PEX for the IO Cells. If that is the only difference, the IOs should be ok.