What are the limitations of the gm/id based analog design methodology employed for sub-micron process nodes? I was reading a research paper that mentioned the design methodology is restricted only to class A devices.
gm/id methodology can be used for any design where transistors are in the saturation region