REVIEW:
What is the intellectual merit of the proposed activity?
The Phase I project of Pragmatic C Software intends to demonstrate
the feasibility of increasing the speed of direct solution in
SPICE circuit simulation. The current solver is extremely accurate but
limited in capacity, i.e. it is slow and uses too much memory. To this
end it is proposed to separate the circuit physics from the computer
implementation and the underlying electrodynamics. The new prototype
computer code will be measured with the aid of commercial circuits.
What are the broader impacts of the proposed activity?
SPICE is the de facto standard for analog circuit verification.
Hence any enhancement will increase productivity of circuit designers
and benefit every company developing electronic circuits.
Summary Statement
The proposed prototype computer code is focusing on enhancing
the speed and capacity of SPICE. In their proposal the team addresses
conceptual questions. However it misses to detail demonstrate the
algorithmic aspects of the proposed software effort and to demo the
proposed effort with examples. If it is simply a question to replace
the equation solver by a more efficient one, then the philosophical
remarks about Newton-Raphson integration rather than iteration are
entirely misplaced.