Practical Chip Design

This morning iSuppli, a highly-regarded industry forecasting operation, slashed their forecast for NAND Flash growth in 2008 from 27 percent to 9 percent. The reason they cited was sharp weakening in consumer spending. This growing weakness appears to be showing up not just in the US but in most major world markets.

In a way this is one of those economic no-brainers. Consumers in the US and some parts of the UK and Europe are on the rack, pulled one way by tightening credit and the other by spiraling food and energy costs. And at least in the US, consumers have so over-extended their personal credit that there's not much stretch left in the old body. That's not good when you are on the rack.

In developing countries, rapidly-rising food prices (take a look at what has been happening in the global rice market), prices for goods and services with high energy content, and government attempts to reduce inflation have all taken deep bites out of consumers' discretionary incomes. The money that might have gone for a new handset with a better camera in it, or for a new iPod (Apple is one of the companies cutting back sharply on NAND purchases) is now going for food and commuting petrol. The fact that iSuppli is still forecasting some growth in NAND orders may be in part that we haven't anticipated the full impact of the global situation on developing-country consumers yet.

For SoC developers caught at the most delicate point in their business cycle—the exposed zone when a project is already underway but has not yet shipped to break-even—this is very bad news. A consumer pullback will certainly mean cancelled orders for existing products, further cost pressure on existing orders (I really should pull this order, but if you can find another $1.50 in your margins somewhere ...), and vanishing interest in any new products that might ship before 2009.

The good news, if you can call it that, is that these slowdowns typically do not hit hard on projects that are just getting started. System OEMs may be sure that they aren't going to announce any new products for Christmas 2008. But they are equally sure that they will be around next year, and that consumers will be back. (This may be a misplaced faith, in the long run, but that would be an economics blog, not a chip-design one.) So they remain interested in really advanced products and in roadmaps.

If you are venture-funded and not expected to show strong revenues this year, or if you are in an organization that can afford to use a long recession to gain an R/D advantage over weaker competitors, this could actually be a good time. Schedule pressure could for once be less important than the pressure to innovate. If, however, you are in an organization that is dependent on current revenues from operations, expect to see the belt get pulled in several notches before things start to look normal again.