What Happens When Ecosystems Stumble?

Economics and design generally are one step removed from each other, but the compression effect of the current design process, combined with extremely low stock prices and a shortage of hard cash are pushing these areas together like never before.

This is new ground, so the ultimate result is somewhat speculative. But given all the factors at play, it appears this downturn coupled with changes in design will have deep and probably long-lasting effects on the semiconductor industry.

Normally, the biggest decisions in system-level design focus on the ability to deliver a competitive system for a reasonable—or in many cases unreasonably low—price in a fixed market window. And while costs increasingly enter into the picture at each process node, those costs generally are well known. Either a company invests in the tools at a new process node and it has enough volume to warrant that investment, or it re-uses an existing process. Even at the leading edge of the technology curve, companies often consider skipping a node altogether because it’s too costly to hit each steppingstone.

Given the rising cost of developing systems on a chip, these are routine business decisions even if they aren’t always pleasant ones. Companies buy IP or they develop it themselves, depending upon similar decisions. And they buy or develop their own tools for the same reasons.

The rising costs are a function of rising complexity, which has forced all companies into doing business in an ecosystem. These ecosystems—and there are a number of them—now stretch from the architectural level all the way out to the verification, debug, test, packaging and manufacturing.

The problem is that in an ecosystem there are weak parts and there are strong parts, and those weak parts can dilute the value of the whole system. Weaknesses are particularly pronounced during downturns, particularly where capital is tight and markets are unstable, because many of the players are small companies.

Unlike in the past, when IDMs would simply move money around from one department or business unit to another in down cycles, small companies in the ecosystem don’t have that option. A second option is to buy up the small companies that are in trouble, but even that isn’t happening now. Acquisitions are rare, despite depressed prices, because there simply isn’t enough credit or capital floating around to make these purchases.

When a critical piece in an ecosystem fails, the whole chain breaks. This isn’t something most companies considered, because it’s cheaper to let startups handle some tasks than to do it in-house. But getting back on track in the event of a failure, or in the event that investments are delayed, could take months if not years. And it could spell the end, at least in the short term, to allowing startups to play critical roles in system-level ecosystems.