Difficulty with Spice Models. Which model to use?

[naeemmaroof]

I am going through MOSIS and I find two different spice models (BSIM4) for 90nm (hopefully for two different specialized processes).
https://www.mosis.com/cgi-bin/cgiwrap/umosis/swp/params/ibm-90/t96w_9sf_9m_lb_3-params.txt and
https://www.mosis.com/cgi-bin/cgiwrap/umosis/swp/params/ibm-90/v15p_9lprf_9lb-params.txt

these have very different (0.26 and 0.12) VTH0.

The problem I have is that I want to replicate the results of certain paper in which they have used 90nm technology. So which model should I use? These two models are only from IBM. So I hope that other foundry parameters will also be much different from these. I found in PTM 90nm model that VTH0 is about 0.38.

What should I do, because for all these models the results replicated are not same (though W/L values are known)

 

[erikl]

There are a lot more parameters than just VTH0 which could be different for processes of the same structure size, but from different fabs/foundries - even if VTH0 is the most important of them, IMHO.

If you don't know the VTH0 of the paper's circuit transistors whose results you want to reproduce, and if the circuit is not too complex, you could try and find their VTH0 through a VTH0-parametrized analysis, or perhaps with an optimization run.