open-source-silicon.dev

Dear <@U016EM8L91B>

I've been extracting netlists from layouts in MAGIC through a command line in Jupyter Notebook.
Nonetheless, something is happening in between when I execute the command line `PDKPATH="${PDK_ROOT}/sky130A" magic -dnull -noconsole mag_files/waffles_nmos.tcl` . I get different parasitics when I execute it in Jupyter or in the main Linux terminal. I don't know the cause of the problem, as I only change the place I'm accessing the terminal. I suppose it may be something to do with the kernel.
I hope you can help me understand what is going on.

Best regards,
Vicente Osorio Rivas

Most likely it has to do with the version of Magic.  The parasitic capacitance modeling was updated relatively recently.  I occasionally update the parasitic coefficients in open_pdks, but I have not done so for SkyWater recently, so I suspect the differences come from the version of magic being used.

There was, however, a very recent change in the PDK where some contact types were missing from a list and so their resistances were not contributing to the parasitics.  I doubt that's what you're seeing, though.