The semiconductor and electronics industries are steadily emerging from the severe downturn of 2001-2002. Although no one wants to be overly conservative while competitors are turning out new products and garnering sales, neither does anyone want to participate in the unrealistic economic practices that preceded the downturn.

The 2001-2002 downturn wasn't just a cyclical slump from normal fluctuations in supply and demand. It was twice as deep and twice as long as any previous downturn, and it signaled a transformation in the electronics industry. Before 2001, the industry was driven by the needs of the enterprise market (i.e., computer technology and communications), and those needs fueled prior recoveries.

This post-downturn recovery will be different: Computer and communications technologies will converge with multimedia technologies driven primarily by consumer demand.

Consumers want a host of features in their products that enterprise markets didn't require-audio, video, global-positioning systems and the like-all in a device that costs $99 or less. They want their systems to be easily portable, too, so designers must be sensitive to power consumption and size.

The complexity dictated by consumer requirements drives semiconductor companies to create chips at ever-smaller geometries. Geometry nodes at 90 nanometers and below require the convergence of previously independent technologies. For example, you can no longer ignore yield issues when designing at the physical level. To meet that requirement, EDA, foundries and integrated device manufacturers must work closely together, giving birth to the field of design-for-manufacturing.

Consumer product chip design also requires the reuse of high-quality intellectual-property blocks. It no longer makes sense to design standards-based connectivity blocks such as USB or PCI Express from scratch when they can be bought off the shelf, predesigned, preverified and standards-compliant. By reusing proven IP, designers can spend their time focusing on their core differentiation. Like DFM, IP directly affects productivity, time-to-market and overall cost of design.

With semiconductor companies developing chips for the much more fragmented consumer market, their profit margins will be lower. Thus their spending patterns will be different. Chip companies will be more cautious, since their profitability will be more precarious. They will require not only the best technological solutions but also solutions that will mitigate cost and risk.

Another important post-downturn shift is the emergence of global markets, particularly in India, China and Eastern Europe. As those economies grow, new consumer markets for semiconductor products will develop.

If consumer demand, silicon complexity, technology convergence and globalization will be the driving forces for semiconductor companies in the coming years, return on investment will be a leading decision maker in the solutions they choose.

Aart de Geus Chairman and Chief Executive Officer, Synopsys Inc., Mountain View, Calif.