A new perspective on the classical microcode compaction problem

Abstract

Some comments are made concerning the generality and complexity of the classical microcode compaction problem. A trivial example is given that no classical compaction algorithm would compact optimally, because the computation of data precedence relations is dependent on the order of the source code. It is further argued that the classical compaction problem can be solved optimally in polynomial time, where the degree of the polynomial is dependent only on the micromachine architecture. We conclude that the most difficult portion of the microcode compaction problem is that of initially ordering the source code.