I am a big fan of design optimization and believe that it is the future… No, no, no, I do not think that all analog designers can be replaced by a tool and Yes, yes, yes I know that the tools have been around a while and that nothing has happened anyway, blah blah.
However, people said that about some early shots on automated flows in the digital world too and hey, they are here now.
Within this research field I plan to search for an optimum optimization tool and learn how to use that in order to relax the requirements on the designer. Also I plan to use the tool to understand relations between vital parameters and performance for complex architectures. For example: what is the area as function of DC gain in a 3-rd order SC filter given a certain specification, etc.
We currently have a project running where we study a two-stage amplifier and how it scales with processes.
There is more to do, and here is a list of suggested final-year projects:
- A script (or some other things) that ties up all the required files from different vendors, such as cadence, mentor, synopsys, tanner, ansyn, muneda, whatever, to run a simulation loop over processes.
- A comparison study of today existing circuit optimization tools
- A theoretical approach to scaling problems