VLSI Processor Architectures
Original Publication Date: 1982-Jun-01
Included in the Prior Art Database: 2005-Nov-11
Software Patent Institute
Philip C. Treleaven: AUTHOR [+3]
AbstractUniversity of Newcastle upon Tyne
THIS DOCUMENT IS AN APPROXIMATE REPRESENTATION OF THE ORIGINAL.
This record contains textual material that is copyright ©; 1982 by the Institute of Electrical and Electronics Engineers, Inc. All rights reserved. Contact the IEEE Computer Society http://www.computer.org/ (714-821-8380) for copies of the complete work that was the source of this textual material and for all use beyond that as a record from the SPI Database.
VLSI Processor Architectures
Philip C. Treleaven,
University of Newcastle upon Tyne
Novel VLSI processor architectures, some implemented by only a few different types of simple cells, are leading the way towards a new generation of computers.
As an illustration of the rapidly increasing complexity of integrated circuits, Charles Seitz of the California Institute of Technology developed the following analogy:
In the mid- 1 960's the complexity of a chip was comparable to the street network of a small town. Most people can navigate such a network by memory without difficulty. Today's microprocessor, using a five-micron technology, is comparable to the entire San Francisco Bay Area. By the time a one-micron technology is solidly in place, designing a chip will be conparable to planning a street artwork covering all of California and Nevada at urban densities. The ultimate onequarter micron technology will likely be capable of producing chips with the intricacy of an urban grid covering the entire North American continent. I
VLSI microprocessors containing 100,000 transistors, such as the recently announced 32-bit microprocessors from Intel (the iAPX 432) and Bell Laboratories (the Mac-32), are becoming more and more commonplaceP'3 In fact, the term "VLSI processor architecture" is normally viewed as being synonymous with such designs.4
But as Gordon Moore, President of Intel Corporation, said at a relatively recent conference, "Beyond memory, I haven't the slightest idea of how to take advantage of VLSI.... How to best make use of the processing technology is what the problem is. "5 The reason for this; pessimism is the escalating cost of designing and testing such complex VLSI processors.
However, Mead and Conway's Introduction to VLSI Systems,6 and the companion Multiproject Chip (MPC)7 courses and silicon foundries8 they helped to establish, have together stimulated the rapid development of a new VLSI proccessor architecture "culture." A brief survey of these novel VLSI processor architectures willillustrate the exciting work going on in the area, most of which is still at the research stage.
A "good" VLSI architecture in the context of this article should have one or more following properties9:
(1) It should be implementable by only a few different types of simple cells.
(2) It should have simple and regular data and control paths so that the cells can be connected by a network with local and regular interconnections. (Long-distance or irregular communication is thus minimized.)
(3) It should use extensive pipe...