The electronics industry has done right by the world in taking an old notion-disaggregation-and pushing it to new heights.
The electronics industry has done right by the world in taking an old notion-disaggregation-and pushing it to new heights. The industry has deployed disaggregation faster and more creatively than any other industry I can think of. It has brought a focus on each of the new subsectors that really is the envy of the world.
After 25 years or so, though, the ride's over, and the next transition to the next business model is not going to be pretty.
It's ironic that the driving force behind disaggregation, Moore's Law, is forcing the model's decay. We've come to a point where the problems associated with design and manufacturing are so acute, and distributed over so many disciplines, that a fractured approach to business can no longer work.
I hosted a panel at the Design Automation Conference that clearly spelled out the challenge ahead. On it were Kaushik Roy from Purdue, Bob Pitts from Texas Instruments, Shekar Borkar of Intel and Jerry Frenkil of Sequence. Each of these men has a different perspective on the problem of leakage, which increases exponentially at 90 nanometers and below.
I would have thought that Borkar or Pitts-hailing from two of the few remaining companies that pretty much cover the waterfront in electronics-would have been in a state of bliss, knowing that internally they could come up with a solution for the industry. Not so. They, the other panelists and the incredibly engaged audience understood that the problem is bigger than any company, and that industrywide cooperation is vital.
The disaggregation model is fraying in other ways. Think about masks. Think about design software. Think about equipment. The mask industry is essentially unprofitable, despite the multimillion-dollar prices; design software falls further and further behind contemporary problems; and the equipment industry spins in its own, masochistic orbit between profitability and tremendous losses.
Don't get me wrong: A return to the era of monolithic companies that own everything from design software to designers to fabs to packaging and assembly is not the answer. But the law of diminishing returns is upon the industry when it comes to specialization.
It's time to start exploring other options.