No End of Killer Apps

It's funny how a certain phrase can get repeated so many times that it suddenly becomes an unquestioned rule. My personal favorite is, "We need to find the next killer app!" In other words, everything would be great if we could only identify the next driver for expensive chips, the next personal computer, and if the industry could go back to the days of high profit margins, wine and roses. Though I think the reality of reduced profit margins and struggling to stay in this business has hit home for most companies, the call for the next killer application is still being heard. I ask myself why.

Well, I have what I think is terrific news. There is no lack of killer applications for semiconductor technology. As has happened in the past, and will continue to happen, the technology itself is not a limitation. We have many examples of this, from the extension of lithography to the subwavelength regime, the increase in transistor drive current from strained silicon, and the enablement of multilevel interconnects with CMP. We are not limited by technology, only by our ability to make it work for us in our lives, where we need it, when we need it and, of course, at the right price.

Take, for instance, power generation. Has there ever been a time when fuel has been more in demand? Do we need cars that consume less gasoline? Yes. Do we need a more efficient way to run our notebook computers so that we can work on them for more than a few hours? Certainly. Could we use backup wireless networks in case of emergency, such as unforeseen natural disasters? That would be great. Fortunately, there are solutions to these needs on the horizon. Technologies such as fuel cells, MEMS, sensors and good old-fashioned ICs are coming to our rescue in these critical areas. These solutions are being worked on and will probably be available within the next few years, miraculously enough. And this is just one application.

Now let's look at health care. Since this is such a broad area, let's take one specific example. Doctors need less invasive methods of diagnosing diseases. Today, there is a pill that can be swallowed, which takes thousands of pictures of the stomach and digestive track. The photos are wirelessly conveyed to a hard drive worn on a belt. Amazing. Could this technology be employed in other areas, particularly in cases where disturbing the part of the body in question with surgery could cause more harm than good? You can think of several, right? Again, semiconductor technology is being utilized for these important purposes. We have many illnesses that haven't yet been cured, but now doctors can better see the problems.

Drug discovery is another key area for innovation. Through modeling and simulation we can test many more drugs before actually using them on subjects, just based on the knowledge we have today about how bodies have responded to the drugs that have been given to them to date. Without a modern computing capability, such detailed explorations wouldn't be possible. We are in the first stages of understanding where drug discovery technology can take us.

Finally, in the area of communication, the applications exist. Just look at what the cellphone has done for our ability to stay in touch, and it's becoming more functional by the day. The functionality, display and convenience of our portable handheld devices keep improving. In the global economy, few things are as important as improved communications.

I could go on, but my message is pretty clear: We don't need a new, huge killer app, because we already have so many. This is the advantage of living in our times, yet also the disadvantage. It is harder to identify which applications to work on first, because timing is so crucial to an innovative technology's success. The overriding challenges lie on the business end, with things like getting there first with the affordable solution. Taking a manageable level of risk on very promising future technology solutions is the delicate balance that is hard to strike.

But I am convinced we have the tools. If any industry has the ability and the wherewithal to rapidly develop intellectual property, it's the semiconductor industry. We may have to borrow techniques from other industries to understand when the timing is right to deliver certain technology solutions, but the needs and demand is there.