Power Predictions 2006: Where Will the Chips Fall?

January 04, 2006 by Jeff Shepard

This is Part Three of a week-long special series looking ahead to 2006.
by Linnea Brush, Senior Research Analyst, Darnell Group

Making predictions about what will happen in the power supply world over the next 12 months is not as exciting as some electronics areas – but much depends on where you look. Power semiconductors and digital power management are two areas (frequently overlapping) that reflect the changes going on in the power supply industry. Trends that began years ago are now gaining traction, and 2006 could signal some critical developments in the industry.

Like it or not, dc-dc converters are slowly evolving away from modular designs to discrete, on-board solutions. The revenue gap between dedicated power supply companies and power semiconductor companies appears to be widening. Both types of manufacturers experienced revenue "peaks" around 2000-2001, and the succeeding economic downturn brought revenue drops. Many companies then began a slow recovery that is still continuing. The difference is that, in some cases, power semiconductor revenue has climbed above its previous peak by about 20%, while power supply revenue has flatlined at about 40% below its peak.

The robust nature of the power semiconductor market underscores both a major shift away from modular power supplies to discrete solutions, as well as semiconductor manufacturers getting into the module business. This trend is not likely to be welcomed by dedicated power supply makers, but it is expected to play a major role over the next year and beyond. The business models of semiconductor companies and power supply companies are different, and the power supply guys simply don't have (or want) the aggressive pricing curves of the semiconductor industry.

Although power supply companies may be facing an uphill competition with chip makers when it comes to traditional solutions, they could find themselves partnering with these same companies to offer digital control of power supplies. Digital power management could emerge as the "breakthrough technology" of 2006, even with its legal wrangling. Indeed, the fact that companies consider it worth fighting over so strongly indicates that these companies have a lot at stake. They believe the market is big enough to justify the expense to defend (or challenge others') digital implementation. The whole thing could backfire, however. Another possible prediction for 2006 is that one of these lawsuits could get dropped or kicked out of court – creating a possible advantage for one or more companies.

The future is far from certain, however. Although many new digital power management and control products were introduced in 2005, customer demand is still uncertain. Combined with the "commercial crossover" of digital power management products, 2006 should also see the first wave of system maker acceptance. Just how enthusiastically OEMs will adopt digital control of power supplies is a wild card. Most of the major computer OEMs are interested, for example, but they want solutions at an equivalent price to existing analog solutions. Selling digital power management as a "premium solution" at a "premium price" won't cut it.

Like discrete semiconductor solutions undercutting existing modular power supplies, existing analog power supplies could undercut digital power supplies. The telecom world is potentially facing this with multiple-output, embedded dc-dc converters that produce lower output power levels and eliminate the need for an intermediate bus architecture.

Price determines the adoption of new technology in the power supply world, and this will be a major challenge for power supply manufacturers in 2006.