Hello All, I am trying to measure Fmax of IBM 90nm technology in cadence. I am measuring S parameters and plotting Gumx. I get Fmax around 500GHz for 90nm while reported Fmax for 90nm are around 300GHz. I am using rfmos transistor & technology uses BSIM4.4 model. I am not sure what I am doing wrong please help.
Probably best to contact IBM, since it's their model.
In reply to Andrew Beckett:
Thanks Andrew. Just want to make sure that if Gumx in spectre is the Masons Unilateral Gain (U) ? Please clarify, I read few contradicting posts
1. Fmax simulation in cadence /blogs/rf/archive/2011/08/11/measuring-fmax-for-mos-transistors.aspx tells you have to plot Gumx to get Fmax
2. Few posts say Gumx is not U http://www.designers-guide.org/Forum/YaBB.pl?num=1272508752/1,http://www.edaboard.com/thread167090.html
In reply to niranjan1984:
Search in cdnshelp for Gumx - you'll see the equation. I'm not really familiar with "Mason's Unilateral Gain"- maybe Tawna will chime in - but the documentation does describe the equation in terms of the s parameters.
Sr. Staff Support AE, Global Customer Support
Cadence Design Systems, Inc.