Intel Delays Its Next Next-Generation 10 Nanometer Chips

The range of computing products powered by leading-edge 14 nanometer process chips is pretty small. As part of Intel's Core-M architecture, they can be found in high-end MacBooks and MacBook Pros, while Samsung's Galaxy S6 line is powered by Exynos system-on-a-chips contstructed with its own 14nm process. More run of the mill high-end equipment is likely to be powered by 22nm process chips (Ivy Bridge i5s and i7s), 32nm process chips (Sandy Bridge i3s and i5s), or even 45nm process chips, which are behind Intel's chips circa 2008, the 2010 Xbox 360 S, and a variety of different AMD processors. Nonetheless, we're already talking about the fabrication process after the next fabrication process, which is the 7nm process IBM recently unveiled as a proof of concept technology.

So, to look at chip process progression over recent years—where "process" is a somewhat vague metric corresponding to the smallest possible distance between RAM memory cells that a given lithographic technology is capable of producing—it's reasonable to think that Moore's Law is right on the money. But there's a big difference between a process that's state of art in the sense of designing system architectures and billion dollar factories around it, and the technology that's just barely poking above the distant horizon. So, it's not all that surprising to hear that Intel is delaying the manufacturing of its next-generation 10nm fabrication process for at least six months until the second half of 2017. The news, as reported first by IEEE Spectrum, comes via a quarterly earnings call from Intel CEO Brian Krzanich. It seems the processor giant has hit a "sticking point."

"Just last quarter we celebrated the 50th anniversary of Moore's Law," Krazanich told investors. "In 1965, when Gordon's paper was first published, he predicted a doubling of transistor density every year for at least the next 10 years. His prediction proved to be right. And in fact, in 1975, looking ahead to the next 10 years, he updated his estimate to a doubling every 24 months. These transitions are a natural part of the history of Moore's Law and are a by-product of the technical challenges of shrinking transistors while ensuring they can be manufactured in high volume."

So, Intel is instead releasing another line of processors based on 14nm nodes, with the upshot being that Intel's Moore's Law based, every-two-ears release schedule—what the company refers to as "tick tock"—is going to take a six month hit. That might not seem like a big deal given that most of us aren't likely to even have a 14nm-based computer in the very near future, but Intel's incredible success over the years is in some large part due to a steady and reliable delivery on Moore's promise. (Recall that Gordon Moore is Intel's co-founder.)

IEEE's Rachel Courtland reiterates what's become a common anxiety in hardware circles: Moore's Law is faltering. Engineering is physics and physics has limits. The universe ultimately becomes indivisible and the closer and closer we get to that point, the more expensive and difficult manufacturing becomes. Worth noting is that Moore's prediction/motivational strategy was largely an economic argument: The more dense an integrated circuit becomes, the costs of making a particular processor component decline. At some point, it has to break down.