ANALISYS: Discontinuous Innovation

By Bob Diefenbacher

The focus of this month's column--discontinuous innovation--attempts to distinguish those innovations that move existing technology forward from those that represent a completely divergent kind of thinking. In this context, the term "discontinuous" refers to the kind of technology breakthrough that cannot be described as an extension or development based on another's invention.

While attending a recent Rensselaer Polytechnic Institute (RPI) event honoring its alumni, I was struck by the use of this term as it applied to the work of a group of distinguished RPI-educated scientists and engineers who had contributed to discontinuous innovation in some way over the past 100 years.

For example, a man by the name of George T. Horton, RPI Class of 1893, conceived of the modern high-capacity elevated storage tank, now common throughout the world. A structure taken for granted today, there were obvious technical issues to be solved when it was first designed.

Closer to our technology world, W. Lincoln Hawkins, RPI Class of 1932, was co-inventor of an anti-oxidant additive that made possible inexpensive plastic insulation of telephone cables, making universal telephone service economical and revolutionizing the communications industry.

One discontinuous innovation that hits mighty close to home, was the result of work done by Marcian E. (Ted) Hoff, RPI Class of 1958. Hoff, otherwise known as the "father of the microprocessor," was the first to recognize that Intel's new silicon technology might make a single-chip central processor possible if a sufficiently simple architecture could be designed. Hoff developed this architecture with just over 2,000 transistors, and invented the first electronic circuit to combine complicated computer functions on a single silicon chip, and set in motion a revolution still underway.

Another noteworthy contributor to discontinuous innovation? C. Sheldon Roberts, RPI Class of 1948, who founded the first company to manufacture integrated circuits, Fairchild Semiconductor. Roberts is perhaps best known for his pioneering research that spawned the microelectronics industry through his work with silicon as a semi-conducting material.

Contrast these findings with the kind of applied research going on in the IT industry today. The innovations currently taking place deal with the production of chips and their physical characteristics. Intel Corporation, for example, recently announced plans to increase its silicon wafer size to 12 inches. This will enable production of 2.4 times as many chips as the current wafers, doubles the surface area of each chip, and decreases the cost of chips by 30 percent.

Intel also says that it will adopt a copper-based technology pioneered by IBM. This will overcome resistance and make chips faster, and chips will be 0.13 micron across, some 25 percent smaller than current ones.

With all the latest buzz about these kinds of advancements in computer building blocks, it seemed appropriate to explore the topic of discontinuous innovation in the IT industry. The truth is, though, I wouldn't classify these recent advancements in chip design as discontinuous innovations. While these developments do represent technology breakthroughs, it would be more appropriate to refer to them as continuous innovations since they really build on existing technology. If technology development in the IT industry is to avoid hitting the wall, however, discontinuous innovation is critical.

One announcement indicates that we have indeed hit a wall in terms of chip performance, opening the door to truly new forms of innovation. Researchers at Bell Labs now say that a size limit has been reached in current materials. The minimum thickness for silicon dioxide film (which does the on/off switching in chips) is 5 atoms, and more likely will stop at 10 atoms. Even 10 atoms is 60 percent thinner than today. Consequently, the road ahead, according to Bell Labs, is paved for the next 10-12 years. After that, we need to jump to a new highway--one in which discontinuous innovation is called for.

We needn't worry that increases in chip speed will vanish over the next decade, though. Indeed, while today's maximum is 600 megahertz; it is predicted that processors in the coming decade will reach 10,000 MHz.

More exciting than that, however, is that the Bell Labs announcement may prompt other companies to come forward with developments they have been keeping secret, according to Richard Doherty, director of research for The Envisioneering Group, a consulting firm that advises companies on trends in the semiconductor industry. Someone may have already discovered the next breakthrough technology in computing power that can put us on a whole new curve.

Computer "smarts" can now be used in many more creative ways making room for new discontinuous innovation in applying computing power to functions and objects previously not considered worthy. Who knows where chips will be applied next? Engineers, do your thing!

The great thing about information technology is that it is populated with talented research and engineering professionals-thousands of them. And all of these great minds constantly ponder and examine theoretical boundaries to find new ways to break through them. Clearly, it is discontinuous innovation that has made our industry today, and will keep it exciting for centuries to come.

After 18 years in marketing and sales at IBM, Bob Diefenbacher founded Denbrook Systems Associates, an IT consulting firm based in Malvern, Pa.