previous | contents | next

of our own invention), and partly because these are the only system of RT modules that are readily available through commercial channels.

In addition, we believe that the emergence of the RT-level is still uncertain enough in its character that it should stay fairly close to its roots. The field is not quite ready for a general treatment of the register-transfer level. In fact, we have in preparation a more extended and scholarly treatment which includes other modules, microprogramming, and switching circuit RT design; the present book can be viewed as an initial attempt to characterize and create this level in an appropriate way.

Although the book uses only a single set of modules (aside from one chapter) and might appear limiting, we urge the reader to go beyond the book and see for himself how general RT level design is carried out. Only some imagination is needed to show that the representation and design techniques are applicable to other RT schemes. We believe the payoff for the reader will be far greater than, say, reading about RT languages for the simulation of hypothetical digital systems.

To those who are interested in microprogramming, RTM's provide a very interesting structure. Microprogrammed and RTM operations appear similar; at each step there are one or more register transfer actions, and input conditions select the next step. There are two substantial differences between RTM and microprogrammed systems: the RTM control part is hardwired (as opposed to being stored in a memory); and the RTM structure can be changed (as opposed to having a fixed data and control part).

A special microprogrammed control module called K(Programmed Control Sequencer) is also. provided as an alternative to hardwiring the control part. Thus, for studying and using microprogramming, its advantages remain, but in addition the very structure can also be changed, providing substantial flexibility.

The book is written in an elementary 'how-to" fashion. Its aim is to get people to design digital systems at the RT level. Our reasons for choosing this style of book can be inferred from the previous paragraphs. Let us note only that a design level exists and becomes real only if lots of designs get created in its terms. Thus a "how-to" book in the area of design plays the same role as does the treatise, with theory and supporting data, in an area of science. This has led us to make various assumptions about our readers.

Thus, we have not discussed number representation in any detail for several reasons. Positional notation for representing binary numbers is being presented in primary and secondary schools. In the rare cases where readers are not familiar with number representation, the concepts can either be "picked up" herein, or be obtained from any book that has anything to do with digital computers or digital systems. Boolean algebra is also not given for the same reasons. Both concepts are not used in substantial quantities, although the reader may feel they are needed to thoroughly understand all of the problems. The concepts of integers and Boolean algebra as given in a programming language (e.g., Fortan, Algol, Basic) are sufficient.

Although RTM's use parts of the PMS and ISP notation of Bell and Newell

previous | contents | next