Due to memory limitations, iterative methods have become the method of
choice for large scale semiconductor device simulation, However, it i
s well known that these methods suffer from reliability problems. The
linear systems that appear in numerical simulation of semiconductor de
vices are notoriously ill conditioned. In order to produce robust algo
rithms for practical problems, careful attention must be given to many
implementation issues. This paper concentrates on strategies for deve
loping robust preconditioners. In addition, effective data structures
and convergence check issues are also discussed. These algorithms are
compared with a standard direct sparse matrix solver on a variety of p
roblems.