Closed JamesTimothyMeech closed 7 months ago
The original data has been lost to time. These are simply what the foundry provided. We did have some of the devices recharacterized, and the data for that is here https://github.com/google/skywater-pdk-sky130-raw-data
No compact model was ever fit to the new data.
Thanks! I will take a look at this!
Am I correct to conclude that this repo: https://github.com/google/skywater-pdk-sky130-raw-data includes one sweep for each device? If we wanted to fit Gaussians to measurements from many different instances of devices of the same type we would have to wait for the measurements to come back from something like this where we can repeat the same measurement many times for many different devices: https://github.com/google/skywater-pdk-sky130-raw-data/tree/main/docs/sky130-testtile-open?
I am running Monte Carlo simulations in ngspice using sky_130_fd_pr devices. When running a Monte Carlo simulation I see that some parameters of the spice models are sampled from a Gaussian distribution. Is the measurement data that was used to fit these Gaussian distributions to the parameters available online? If so where can I access these raw measurements?