Panel examines disruptive technology
By Ron Wilson
SANTA CLARA, Calif. — The concept of disruptive technology and its potential for predicting winners and losers in the semiconductor industry was explored by a panel of business and technology gurus Wednesday (Jan. 23).
The panel was led by Harvard Business School professor Clayton Christensen, author of "The Innovator's Dilemma," the book that introduced the concept of disruptive technology. Christensen summarized his concept: when the technology serving a market begins to exceed the needs of its users, there is a potential for newer, less expensive, less capable technology to displace the original market driver. The displacement of minicomputers by the personal computer illustrates this concept, he said.
Before a disruptive technology appears, the market leaders — typically highly integrated companies, as was the case with Digital Equipment Corp. in the minicomputer world — produce proprietary, highly optimized solutions, which tend to have high profit margins, Christensen said.
But disruptive technologies do not compete on performance, and hence tend to be modular, organized around de facto architectural standards, and provided by a disaggregated supply chain, in which the best margins usually go to the module suppliers, not the OEM. Such technologies also tend to provide utility to a much wider portion of the market than the optimized technologies they displace, Christensen said.
Other panelists — Jonathan Joseph, managing director at Solomon Smith Barney; Richard Newton, dean of the College of Engineering at UC Berkeley; Nick Tredennick, editor of the Dynamic Silicon newsletter; Chris Rowen, president and chief executive officer of event co-sponsor Tensilica Inc.; and EE Times editor-in-chief Brian Fuller — examined this model from the point of view of technology challenges, either in architecture, the ability of customers to use the technology or its ability to predict winning companies in the semiconductor industry.
In an additional presentation, Rowen argued that his company's configurable CPU cores fit the model of disruptive technology, offering a substantial cost/performance advantage over general-purpose microprocessors while opening the realm of CPU design to a wider range of designers.
The fit was not coincidental. Christensen has been a consultant to Tensilica for some time, and is close to Rowen. "Clayton's ideas have become something of a gospel around here," one Tensilica manager commented after the event.
In a question and answer session, Fuller asked if system-on-chip technology was itself disruptive. Christensen suggested that it was, to some vendors such as Intel Corp., but that it could in fact sustain the market positions of other companies. "What's sustaining to some is disruptive to others. It is a very relative concept," he explained.
The author also opined that wireless handheld devices could prove disruptive to personal computers, and that his model suggested that semiconductor processing could shift from the present focus on large batches of wafers to very fast throughput of individual wafers. When asked about data supporting his conjectures, Christensen warned that, in the case of Digital Equipment, the disruption had already become irreversible before market data indicated that it had begun. "In times of disruption, a good theory gives a better view of the future than current data," he said.
Ron Wilson is editorial director of Integrated System Design, a sister publication of EE Times. |
