Hi @Tim Edwards, if my repo has a lot of designs and some of them should be compliant to a specific set of requirements. Do I have to create a specific "cace" directory for each of them and fill it with tests? How can I select which one evaluate? The only "requirement" to use cace is to have this "cace" directory with tests and the .txt, right? The

sky130_ef_ip__instramp

design fails on some tests. Is that expected?

Unless you want to separately characterize sub-components of your design, I suggest that you just have one directory for running cace on the project's top level. However, if you do have multiple characterizations, you just need an additional characterization file. If you run

cace_gui.py

without arguments, it will look for a

.txt

file with the same name as the project directory. If you want to point it to a different file, then just pass the other file as an argument. I think the only requirement is that within the

.txt

file, the

paths

declarations need to point to valid directories. Yes, the

sky130_ef_ip__instramp

is a barely credible (if that) design and will fail a lot of specs (I more or less fused together a random schematic with a random set of specs).

Just to be sure, it's posible to: • Have multiple sets of testbench per module, existing on different folders • And store all the

.txt

files on the same directory below root. The only requirement is that the

.txt

file indicates the correct

.sch

folder, according to which module is evaluating. Is that right?

@aquiles viza: I think so. Please note that I have not worked through many use cases. I'm happy to try to accommodate any new use cases that don't break the existing methodology.