Increasingly, we are hearing about consolidation in the semiconductor industry. Bill McClean of IC Insights has perhaps pointed this out most clearly, noting that the top 10 capacity leaders held about 60% of the total 2009 worldwide IC capacity, and the top 15% held a 71% share. Looking at only 300mm capacity, the consolidation is even more obvious: the top 10 leaders held an imposing 84% share with the top 15 companies holding all but 8% of the world’s 300mm IC fabrication capacity, according to McClean.
This is good news for those15 companies, but not such good news for consumers. "With so few companies representing such a large share of the leading-edge IC production, IC Insights’ long-held belief that the companies left standing after the ‘shakeout’ will reap the rewards of increased profitability, is now coming true. However, for the IC user, this also means that IC average selling prices are unlikely to decline as they have in the past. The pricing pendulum is now swinging in favor of the IC producers and it may not be swinging back for a long time," McClean said.
With a greater percentage of spending coming from a shrinking number of companies, semiconductor industry capital spending is becoming more concentrated. Because of this trend, IC industry capacity is also becoming more "concentrated." As tight as overall IC capacity has been in the IC industry since the first half of 2009, 300mm capacity has been even tighter, McClean notes. In 1Q10, 300mm capacity utilization was 97%, which is essentially indicate a "sold out" situation.
With most capital spending in 2010 going toward developing and moving to finer feature sizes (with little going toward adding wafer starts), the availability — or lack of availability — of IC devices will become a major factor in the second half of this year. Ion fact, IC Insights believes that one limiting factor to 30% or greater IC market growth in 2010 might be the lack of IC fabrication capacity needed to support such growth.
Also expect higher average selling prices (ASPs) for almost all types of ICs. "Given that almost all DRAM, NAND flash memory, and microprocessors are produced using 300mm wafers, it follows that each of these segments registered increasing average selling prices throughout 2009 and into 2010," McClean said. "IC Insights believes that buyers of these IC devices should be prepared for similar ASP trends for these products in the second half of this year."
While all of this is going on – the consolidation, sold-out capacity and higher selling prices for mainstream ICs – I keep thinking applications that fall outside of the mainstream. Where are those apps that require different types of equipment and materials technology and different process technology? Do they exist? How big of a market do they represent? Might they be a way for suppliers to diversify outside of the mainstream 15-companies-now-controlling-all-the-spending semiconductor industry?
My list of applications that offer a potential way to diversify for companies offering thin film deposition and patterning capabilities include: MEMS, hard disks and read-write heads, flat panel displays, optical disks, photovoltaics, LEDs, CMOS image sensors, and superconductors. The MEMS market is perhaps the most intriguing of these since it encompasses so many different types of devices, from microphones to sensors to micro-fluidic labs on a chip.
But perhaps the application with most explosive potential is intelligent medicine. Imagine taking a digestable "smart" pill that is in essence an integrated circuit. As that pill works it’s way through your body, it monitors the status of all kinds of functions and reports the results to your iphone or similar device. That kind of technology is, in fact, already here. Proteus BioBed in Redwood City, CA has developed what it call the Raisin™ System (there was an Apple on your desk, a Blackberry on your belt and now a raisin inside, quips Proteus CEO Andrew Thompson). The initial application of the Raisin System is for the treatment of patients with heart failure. The system senses and records the precise time a patient takes one or more microchip-enabled drugs, providing physiologic feedback and decision-support to the patient, caregivers and clinicians, thus facilitating a cost-effective pathway to improved patient outcomes through personalized medicine.
How cool is that!? It doesn’t take much imagination to envision all kinds of medical apps with new and unique requirements that could leverage traditional IC manufacturing technology. This market could be huge, given the aging baby boomer population and over-loaded healthcare system. Consolidation? Who cares? Bring on the new apps!