We study different memory cell designs and compare their advantages an
d disadvantages from an engineering point of view. We look at operatio
nal stability as a function of temperature and stray charge (random ba
ckground charge), and discuss the issue of reliable mass production. W
e conclude that memories seem to be one of the most promising large sc
ale single-electron tunnel applications, that lie, particularly when g
ranular films are used, already in the range of today's process techno
logy.