|Reference||STATE OF THE ART||Stan Augarten|
|The First IC Made with Computer-Aided Design|
In the early days of the IC - the 1960s to mid 1970s - the designing of chips was an undefined art, the creative effort of an engineer or team of engineers sitting at a drawing board and making up the rules as they went along. There were no textbooks on IC design and no college courses on the subject. Computers weren't widely used in the design process, both because they were still very expensive and finicky machines, and because ICs were not yet so complicated as to be beyond the ken of their designers.
Computers made their first significant foray into IC design in the mid-1960s, when Fairchild, the matriarch of the semiconductor industry, began experimenting with computer-aided design (CAD). CAD allowed Fairchild to make more complicated chips faster and more reliably; today's fantastically sophisticated ICs, with their hundreds of thousands of microscopic components, would be well-nigh impossible without the use of CAD systems, now a small industry in their own right.
Fairchild's first large-scale CAD chip was called Micromosaic. (A large-scale IC is one with up to 100,000 components.) It was a logic chip with about 150 AND, OR and NOT gates (p. vii). The basic chip consisted of a few hundred transistors, which could be hooked up in almost any pattern by rearranging the chip's aluminum interconnections, in an electronic version of a connect-the-dots game. Some transistors were bypassed, others connected. Determining exactly what pattern of gates was called for was the job of Fairchild's computer, which made such decisions in accordance with the customer's specifications.
Thus the chip could be optimized for specific applications, such as computers designed chiefly to solve scientific problems. Since the deposition of the aluminum interconnections is one of the last steps in the IC fabrication process, customizing the Micromosaic was not especially expensive. The initial layers of the chip were made in the usual fashion (p. xi), and only the final photomask, the one laying down the aluminum, had to be custom-made.
Before the advent of the Micromosaic, ICs were entirely hardwired - that is, they consisted of predetermined arrangements of logic gates whose layouts could not be changes. These layouts were not necessarily best for all applications, but hardwired chips were the only ones on the market. Customized chips, like the one on the right, alleviated the problem somewhat. However, it wasn't until the invention of the microprocessor (p. 30), with logic circuits that could be programmed to execute almost any operation, that ICs became truly flexible.
|The horizontal light beige lines are aluminum interconnections, which may be arranged in almost any configuration to satisfy the user's needs. As a result, one basic chip may be modified to serve many purposes, with the resulting economies for Fairchild and its customers. the dark vertical lines are bipolar transistors. Actual size: 0.150 inches square.|
|STATE OF THE ART
©Copyright Stan Augarten
|This book is provided for general reference. The National Museum of American History and the Smithsonian Institution make no claims as to the accuracy or completeness of this work.|
|page:||Index 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76|