 |
July 26, 2005 — Since its launch into the marketplace nearly three decades ago, MEMS has enjoyed growth rates on the order of 20 percent per year. That is hardly surprising for newly emerging devices at the forefront of their commercialization. However, it appears that MEMS has reached the point where, on a revenue basis, annual growth rates are expected to start dropping into the single digits.
That doesn’t mean that the opportunity for MEMS is diminishing. Quite the contrary. Unit shipments are expected to continue to see double-digit growth rates for the foreseeable future. The explanation for this change is very simple: price.
The price of MEMS devices has dropped to the point that it no longer has the impact on revenues that it once did. And with the next generation of MEMS devices entering the market at near commodity-level pricing, their ability to push revenue growth will be virtually nonexistent. One hundred million (or more) of anything, with a price of $1 (or far less), has little to no impact on a multi-billion dollar market.
Here’s what’s happening. Commercialization of MEMS devices clearly occurs in waves. The last wave of devices to enter the market was gyros and optical MEMS. Both were very expensive compared to accelerometers, pressure sensors and inkjet nozzles, which comprise the majority of MEMS revenues. As a result, having millions of devices entering the market with prices that are 10 to 100 times greater than the average selling price of MEMS overall, is certain to drive revenue growth — at least for the short term. At some point, a crossover takes place, where the emphasis on growth switches from the reliance on high prices (due to low initial volumes) to unit volumes (at significantly lower prices).
It appears that gyros and optical MEMS are the last of their kind; that is, devices that enter the market at very high prices. The next wave of MEMS devices poised to shake up growth is entering the market in a completely different way. In the near term, those include RF MEMS (which are already making a mark) and microphones. Both are devices that will rely on extraordinarily high levels of unit shipments to drive revenue growth. But while those revenues will be good for those companies on an individual basis, they will do little to actually drive revenue growth for MEMS as a whole. Rather, their impact will come in the form of keeping unit shipment growth high.
This is evident in my ranking of the Top 25 MEMS suppliers (based on 2004 revenues and unit shipments). In terms of revenues, Hewlett-Packard was displaced as the No. 1 supplier for the first time by Texas Instruments. TI achieved this because the average price of its Digital Light Processing systems far surpasses that of HP’s ink jets, despite the low volumes that TI ships compared to HP. In terms of unit shipments, HP remains solidly in first place.
Conversely, Agilent Technologies jumped from No. 15 to No. 6 in terms of unit shipments, and is expected to be a strong contender for the top three next year. And even though it cracked the top 10 for revenue (up from No. 20 last year), it’s not likely that it can move up much further. Those at the top of the rankings are generating revenues that would be virtually impossible for Agilent to achieve, given the low price levels of its products.
Looking a little further out, this model of growth driven by units rather than revenues will continue to be sustained, as evidenced by the next wave of MEMS devices in development. For lack of a better term, I’m referring to this cluster of devices as “transparent MEMS.” These are devices that will be extremely difficult, if not impossible, to track as individual products from a market research perspective, as the intent is to integrate them within the electronic structure of microprocessors, and in some instances, the silicon wafer itself.
As such, the door has opened to an entirely new class (and use) of MEMS, and the unit shipment possibilities are staggering. This is perhaps the best indicator of the gradual shift from the revenue-driven growth MEMS has experienced to date, to the unit-driven growth model of the future.
On a final note, early last year I predicted what has proven to be a significant development in 2004: It was indeed the year of the acquisition. Fourteen companies were acquired for just shy of $1 billion, with the average deal valued at roughly $70 million. Even better, a number of those acquisitions were made by MEMS startups, indicating a nice maturing within the broader market.