René Penning de Vries, senior vice president and chief technology officer of NXP Semiconductors, spoke to EDN about how design and R&D are changing as we evolve past the traditional definitions of Moore’s Law and into a new era based on ＆#118alue-added applications and guided more than ever before by economics. Excerpts of that interview follow.
Where do we as an industry stand with Moore’s Law? Are we coming to its end?
No, I don’t see that at all. I think Moore’s Law is going to have a future but only for the applications that have such a high volume that the investment associated with the next generations can be earned back. Saying that limits the application to very few that we all know about and thereby [limits Moore’s Law to] very few players that we all know about. Of course, it’s about the memories, it’s about the big processors, and it may be about the basebands in wireless handsets. Those are the applications that can create the accumulative business that justifies the enormous investment. More and more, companies must find for themselves a place where they can make business, and, in some cases, that [idea] leads to the conclusion that that [place] is not Moore’s Law or not an application based on Moore’s Law.
So applications become the next industry driver.
Yes. What you will see more and more is that the big concerns that the world is facing at this moment—the need for sustainable energy, the need for water, the need to reduce the cost of medical care—those challenges can only be realized as more and more we create intelligent solutions. Intelligent solutions in my view means optimized solutions for a particular application area. That [idea] is not necessarily based on Moore’s Law technologies. On the contrary, often that [idea] will ask for technologies that can deal with high voltages, technologies that can deal with special requirements. Typically, these are technologies that are based on a relatively old CMOS generation but with a sophisticated device or a sophisticated addition on top of it that makes this specialty an option. … Technically, I do not believe Moore’s Law is over. It is economics. There is a famous analogy in the industry that relates to commercial airplanes. The Concorde flew higher than the speed of light between Paris and New York. Technically, it was a very feasible proposition. Economically, it didn’t work out. It was just too expensive. That’s why, for commercial applications, we don’t see the Concorde anymore. The development in itself did not limit evolution in the airplane industry. We saw airplanes become more efficient, less noisy, less polluting, more pleasant to travel in. There has been a continued line of innovation, not along the lines of making it faster, but along the lines of having a smaller footprint.
Considering the economy and this evolution to a “more than Moore” era, where are we in R&D and spending on R&D?
It will change in nature. Whereas we had massive amounts of money spent in developing new technology, we now see the wealth of applications coming toward us in dedicated solutions. It will move from the foundation process technology into the area of application-specific, customer-optimized solutions. … We’ll shift more toward customers, toward applications and move away from very deep process technology.
Will it be more difficult to innovate with reduced R&D investments and smaller staffs? There have been massive layoffs in the industry over the last year.
For sure, it is a stretch. But we also know that, after the crisis, there is going to be an upturn again. To me, it’s a matter of endurance. We have to sustain innovation. We have to continue to invest people, money, and brains in creating innovations that later on we can harvest.
