The Harder it Gets for Chipmakers, the Brighter the Future for Printed Electronics

It’s an economic reality, which engineers hate to concede, that in the marketplace the better mousetrap seldom wins. It is awfully hard to displace an accepted technology, which has in place a huge infrastructure, immense knowledge-base and has been proven to work, with a new technology that promises to do a better job. Just ask those folks who have gone up against petroleum-based technology in the internal combustion engine or (closer to home) bulk silicon in the semiconductor manufacturing process—being better isn’t enough.

The fact is that large, established industries, rarely embrace truly innovative technologies unless they are forced. Often even the promise of new applications and markets is not enough. There is too much cost, work and risk in bringing a promising idea from a laboratory to a volume-production reality. But this doesn’t mean that the future isn’t bright for printed electronics. Today, the laws of science and economics are conspiring to force the technology industry to embrace new production processes. In fact, it will be the demise of Moore’s Law that gives life to the printed electronics industry. Pardon the pun, but the writing is already on the substrate.

For much of its life, Moore’s Law was driven by design shrinks on increasingly large bulk silicon wafers—processing power increased, transistor costs dropped and new applications arose. Let’s face it, as long as that trend continued there wasn’t a big need for new ways of doing things. But things are changing. Device shrinks can’t get it done alone anymore. Expensive processes that leverage new materials, like SOI, strained silicon and hi-k dielectrics are necessary. Moving to larger wafer diameters is also getting more costly and seems to be at a point of diminishing returns.

Even as Intel and IBM announced in late January of this year that by leveraging new materials—on top of new design architectures, which use more, rather than less, of that valuable wafer real-estate—that they were going to move to the 45-nm node, Texas Instruments stated that the 45-nm node would be the last for its production business. TI is not alone. Cost has become such a big issue that many semiconductor companies have not only gotten out of manufacturing, but are lingering longer at designs around mature processes, as in many markets cost becomes more important than raw processing horsepower.

Trends that will Drive the Move to Printed Electronics
There are several big drivers for a printed electronics future: (1) as process costs skyrocket and markets are increasingly made up of price-conscious consumers, there is growing demand for low-cost production processes that do not involve bleeding-edge capabilities; (2) in addition to cost concerns, form factor is a big focus in large consumer markets; (3) the technology industry is already making huge investments in learning to design and manufacture around new materials, and this will only accelerate; (4) alternatives are increasingly being sought to the conventional CMOS, and as Moore’s Law comes under even more pressure the need for alternative technologies will grow.