Hello everyone, I am trying to extract a layout to spice using the commands 'extract all' and then 'ext2spice', however I am getting w= and l= values which are 200million times larger than they should be. For example a 0.15 micron gate length extracts to l=30. Please can you let me know whats happening and how I can fix it? I have tried running the commands "ext2spice scale off" and "ext2spice hierarchy on", which makes l=150000u (but still wrong.)

These values are correct. There is a

.option scale

line inside the model parameter files for the process that means that all distance values must be in units of microns (and all area values in units of microns squared). It is quite unorthodox, and I made some effort to convert it back to SI units, but it is a tricky problem and I would have to have a very thorough CI set of test simulations to make sure I got every single equation in the models correct. So I have not done it, and the units are still in microns.

Thank you Tim, I'll try that!

Also ask John how interested he is in either Global Foundries GF180MCU or SkyWater 90 FDSOI. Both are upcoming (GF before the end of the year, Sky90 sometime next year).

Aloha. July 2023. Status of SkyWater 90? Mahalo

The future of an open sky90 is highly uncertain right now, tending toward the doubtful.

Mahalo. Earth ICT, SPC is preparing CHIPS new construction fab final application. New fab focused on optimize Oscillator for Adiabatic Computational Circuitry. Michael P Frank and Robert W Brocato new June 6th grant OACC patent https://twitter.com/MikePFrank/status/1669086679804649472?s=20 Earth ICT, SPC focus extremescale, sci and ai training. DOE ACE program. Google, MS, AWS, Tesla custom/bespoke AI training si. If no new fab ... plan b ... modify enhance existing ... plan c use what's available, e.g. 180, 130, ... 90 available just no as efabless MPW... possibly some 28nm fab for a price. Beaucoup Mahalo for your, and efabless FOSSI community impactful work.