Exclusive Research: Where Are The Midsize Companies?

Data shows plenty of large and small companies, but not much in the middle.


By Ed Sperling

In the chip world it’s beginning to look like the erosion of the middle class, or at least the middle tier.

The bulk of companies working in the semiconductor industry—from design to development—have annual revenues of either less than $10 million (26.2%) or more than $500 million (25.4%). The number between $10 million and $100 million is only 12.8%, while the number between $100 million and $500 million is only 8.8 percent.

These numbers reflect several important trends in the chip design world. First, there are far fewer IPOs than in the past, making the exit strategy for investors less clear cut. That, in turn, limits the number of new companies being started, particularly in the United States, Europe and Japan.

Second, companies that do achieve enough mass to hit the radar screen—usually with successful products and unique technologies—often become acquisition targets for larger companies. These acquisitions are not always amicable, though, which has made integration and assimilation much more problematic than simply acquiring a technology or a technology team and developing technology in-house.

Third, the table stakes for getting into the system-level design world are much higher than in the past. Technology is more complex, and often it is the integration of more than one piece that makes it attractive to potential customers. While chip chip design and development always seems complex compared with what came before—everything from lithography to substrate materials, packaging, gate design and tradeoffs between power and performance—solving these issues can no longer be done serially. Architects can no longer just pass the concept to the designers, who in turn pass it along to the developers and verification engineers, and then pass it along to the software coders.

Finally, despite the perpetual development of new technology, the chip design world is maturing. That is marked by relatively stable single-digit growth by companies in good times, and fewer entrants into a market. While the industry is still viable, with enormous opportunities for players in the industry, the number of companies entering the field will be limited until there is a radical shift in technology, which usually happens when something is broken or there is a huge demand for new technology.

The data also shows that the bulk of the pre-design activity was done by the design engineers, who accounted for 44.6% of all work. Engineering and executive management entered into 20.1% of all decisions, a reflection of how much money it now costs to develop a new chip.