Engineering and Management Principles of Firmware,

A computer system consists of three distinguishable parts hardware, software, and documentation. The question is where does firmware fit into this scheme. In terms of logistic support the best way to think of firmware is that it is software in a can. The can or medium may be a single integrated circuit or a printed circuit card full of integrated circuits or even an array of printed circuit cards. Whatever the medium, the fact remains that canned software will have hardware characteristics, and its contents will have software characteristics. The dilemma presented by using firmware has an economic basis. The advent of microprocessors has catapulted firmware into being a very economical means for a system manufacturer to house his software. Now, he needs not worry about having a means to load his programs, and what might happen to them if his system looses power. He avoids the cost of some expensive peripherals, complicated primary memory schemes and backup power sources. He can now sell the package with the software completely embedded and invisible to the customer and a potential competitor. He has, however, given up the capability to make easy changes to the system. This is of minor concern in a static, exhaustively tested, system.