An ad campaign a few years ago posed a clear strategy choice: Whether to be wolf or sheep? Most semiconductor companies face a similar decision today: be active in market consolidation or risk being swallowed. Portfolios of half-empty fabs burning cash reduce strategic options and can threaten an IDM's existence.

Many consolidation strategies are bound to fail. One example is entering the foundry business when the number of underutilized fabs is growing and overall demand is absent. Succeeding in the foundry business requires far more than simply offering fab capacity. Transferring products to other fabs takes too long to achieve a significant impact, if begun under pressure.

Does it makes sense to migrate older products to more advanced technology levels, especially if fabs require upgrading? Attempts to close fabs operating below break-even often fail due to long-term customer commitments for existing products.

These strategies have one thing in common: poor timing. Companies typically wait too long " until drastic measures are needed " instead of identifying sustainable, long-term options early on. If a fab lite strategy or fabless business models are inevitable in the long run, IDMs are better off planning well in advance to keep their options open.

It's time to shift the thought paradigm from "What fabs do I have today, and how do I fill them?" to "What fab base do I need five years from now and how do I get there?"

Foundries can support that transition in multiple ways: By transferring legacy products into their fabs to enable earlier fab closures; migrating products to newer technology nodes not available through IDM fabs; or by assuming fab ownership early on to convert it into a foundry fab, and loading additional foundry business over time.

Such conversions take time. Starting early is a critical factor for success since empty fabs often don't sell. Selecting the right foundry partner for proactive fab base consolidation is vital. A long-term partnership with a financially stable foundry is better than the lowest wafer prices in the short term. The reason is that IDMs, despite active consolidation efforts, could fall victim to foundry market consolidation if their low-cost supplier fails to survive the current downturn.

Therefore, starting consolidation early is essential to avoid making decisions under pressure or limiting options. IDMs should stress the reasonable life expectancy of the potential foundry partner over price.

Finally, IDMs surviving the current downturn should not wait for the next slump--they should act now, not react.

—Norbert Meyer is X-Fab's vice president of strategic projects.