Bell Labs Won't Have to Share Ideas as Much

Communications News

September 1, 1983

Bell Telephone Laboratories has been the premiere research and development organization in the world, singly responsible for literally thousands of inventions and discoveries that have changed both the world of telecommunications and the "outside' world in general . . . with its most often-cited accomplishment being the transistor.

Because of its affiliation with a regulated monopoly and the 1956 consent decree, which required Bell Labs to make available to all others licenses for its patents available at a nominal fee, many of the wonders flowing from the Labs couldn't be fully exploited by the Bell System. As Continental Telecom Chairman Charles Wohlstetter observes: "With all of the technological expertise in thousands of American companies, it was Bell Labs that invented the transistor, and thus made possible the computer. And for its discovery, Bell was compensated by being prevented from entering the computer business.' So who said life was fair?

But Bell Labs certainly has received its share of praise for the many accomplishments. Among only a few of its pioneering achievements--which could fill a book, and have--are a demonstration of the first high-fidelity sound recording in 1925, origination of sound motion pictures in 1926, the first long distance TV transmission in 1927, the transistor in 1947 (for which its inventors received the 1956 Nobel Prize for physics), the laser in 1958, magnetic bubble devices in 1966 . . . and on and on.

Bell Labs was incorporated in 1925, largely from research activities of Western Electric that dated back to 1907 . . . although actually the lineage could be stretched all the way back to Alexander Graham Bell's laboratory in 1876.

During 1982, more than 25,000 worked for the Labs at 21 major locations, with some 30 percent of them assigned to lab locations at seven Western Electric plants. And what a staff! About 3,300 of them have PhDs and another 6,000 have Master's degrees. Some of them have since left, of course. About 4,000 were transferred to American Bell and another 3,000 are destined for the Central Services Organization serving the regional operating companies.

Reflecting the increasing emphasis on computerization of telecommunications systems and other areas, more and more Bell Labs people are working on software-related projects. For some hard numbers on software activities, nearly half of all Bell Labs employees are working in the area of software development or software support, whereas only 15 percent were nine years ago. Bell Labs is also a large user of computer systems, with 1.5 computer terminals per technical employee and something like 1,800 host computers scattered around its facilities. According to Vice President Eric Sumner, "We support about 35 million lines of live code in the Bell System. That probably makes in the biggest software enterprise in the world. We are in software in a big way.'

The computer industry has high hopes for the Bell Labs-developed UNIX operating system, and AT&T has begun commercial marketing of a new version of UNIX, which is expected to become a major force in newer generation computers. (Also see the section on Western Electric for more on UNIX.)

Production of the Bell Labs Bellmac-32 microprocessor began last year. Containing 150,000 transistors on a chip smaller than a dime, it will be popping up in a wide assortment of equipment . . . including a forthcoming line of small computers, expected to be marketed by AT&T Information Systems.

Today, even with a scorecard, it's virtually impossible to keep up with the fastchanging developments in integrated circuits. Even Bell Labs has to be amazed at what's happened to its little device in a little over 35 years. Trying to put it into perspective, Bell Labs President Ian Ross explains, "In each of the last 10 years we have doubled the number of transistors we can squeeze on a chip of silicon. There are over 600,000 components on a 256k random access memory. Compared to the cost of a quality transistor 20 years ago-- about $1 to $10--the equivalent costs of a transistor in a chip today is something like one-hundredth of a cent--a thousandfold cheaper. If we had had the same progress in the aircraft industry, you and I could be flying between London and New York in 500,000-passenger planes, and the fare would be only about 25 cents!'

But how much longer can this trend continue? According to Ross, "The progress in microelectronics will continue for some time. One-megabit chips are being designed today. We should see one working in 1983. Four-megabit chips are a possibility by the late 1980s. Metal-oxidesilicon (MOS) circuits, using features of less than a micron in size, have operated 100 times faster than is normal in today's production devices, with logic circuits switching in less than 10(-10) seconds. Gallium arsenide semiconductors and Josephson junctions are other possibilities for ultra-speed devices.'

Lightwave technology has been another of Bell Labs' great-achievement areas. Bell Labs and Western Electric engineers have already tested a long distance lightwave system that can carry more than 100,000 two-way telephone conversations simultaneously over a glass fiber about a tenth the thickness of a human hair.

Earlier this year, Bell Labs demonstrated the first practical communications laser whose output can be tuned electronically from one ultra-pure single frequency to another, which Bell Labs sees having dramatic implications for future generations of lightwave communications systems. Called the "cleaved coupled-cavity' (C3) laser, the semiconductor device has properties for improvements in lightwave system capacities and longer unboosted transmission distances. It's already been used in a transmission experiment that set a world record. Unboosted signals generated by the laser pulsing 420 million times per second traveled about 75 miles, error-free.

Arno Penzias, vice president of research at Bell Labs, recalling the introduction of the first standard-order shortdistance lightwave system in Atlanta in 1980, says, "We now have such systems in over 50 cities. Together they total more miles of lightguide fiber than has been installed in all the rest of the world combined.' This past February, a major segment of AT&T's Northeast Corridor lightwave system was placed in service, as explained elsewhere in this report.

Some have been concerned that one of the casualties of the divestiture would be the famed "blue-sky' research conducted by Bell Labs, now that its parent is entering a more competitive world without the safety net of the operating companies. AT&T is on record, however, saying it is committed to continuing "an extensive program of basic research.'

Next year, meanwhile, Bell Labs will continue to provide the divested companies with technical information they request. And until 1987 it will be giving them assistance in providing the required interconnections to interexchange carners.

To the outside world's loss, however, Bell Labs will no longer be required to offer licenses to others for non-BOC items. Hopefully, that won't wind up being one of the dearest prices paid for the divestiture.

Photo: Bell Labs has made startling advances in semiconductor technology since it invented the transistor in 1947. This digital signal processor chip in a Number 5 ESS, for example, is smaller than a Touch-Tone phone button, yet it's packed with 45,000 transistors, able to make a million calculations per second.

Photo: When completed early next year, AT&T's 776-mile Northeast Corridor lightwave project will stretch from Virginia to Massachusetts, carrying huge volumes of calls.