Back in 2011, former Intel (NASDAQ:INTC) CEO Paul Otellini teased on a conference call that it would be making a game-changing announcement. That announcement, which came in May 2011, turned out to be that Intel would be implementing FinFET transistors in its 22-nanometer manufacturing technology.
FinFETs allowed Intel to get a significant jump on the competition, delivering what microprocessor expert David Kanter suggested was a two-generation leap in transistor performance. Intel began mass production of its 22-nanometer technology in late 2011, and the first products based on this technology -- known as Ivy Bridge -- hit the shelves in early 2012.
According to Kanter, Intel may be on the verge of another FinFET-like breakthrough at the 10-nanometer technology node.
Kanter, after spending considerable effort following Intel's published research, suggests that for Intel's 10-nanometer chips, the chip giant will implement what are known as Quantum Well (Field Effect Transistors) or QWFETs.
This technology, according to Kanter, should allow for "excellent transistor performance at nominal operating voltages around 0.5V." This, says Kanter, is compared to "roughly 0.7V for FinFETs."
In other words, this technology should allow Intel to deliver faster, lower power chips.
Intel adopts at 10-nanometer, while the rest of the industry follows at 7-nanometer
Kanter believes that it is very likely "given the timing of the research and Intel's historical cadence of new manufacturing technologies," that Intel will adopt this technology at the 10-nanometer node rather than wait for the 7-nanometer node.
Kanter also thinks that Intel's competition will "wait until the 7nm node to use [Quantum Well Field Effect Transistors]."
Could this be why Intel is so quiet on 10-nanometers?
Intel has been extremely quiet about the timing and the details of its 10-nanometer manufacturing technology. The company has said that it plans to continue to very aggressively scale the size of its transistors and interconnects at the 10-nanometer node in order to deliver the typical generational cost-per-transistor benefits.
If Intel is planning on doing such aggressive chip area scaling and moving to a radically new technology, then it's no wonder that the company is being silent on the timing of its 10-nanometer ramp. Getting a technology with so many improvements on so many fronts into high volume manufacturing is likely incredibly difficult with a lot of uncertainty involved.
Is this Intel's silver bullet?
If Intel can get 10-nanometer into volume production by the end of 2015/early 2016 as Kanter suggests, then Intel should be able to maintain, if not extend, the manufacturing technology lead that it has over its foundry peers.
The good news is that if Intel really does have a "revolutionary" 10-nanometer technology on its hands, then this could help its mobile ambitions -- as long as Intel's product definition and implementations can execute.
That said, Intel had a similar lead with FinFETs for quite a while over the competition, but it failed to capitalize on it as far as shipping mobile products go. Having a manufacturing technology lead is certainly an enabler, but product execution is, in my view, more important.
As an Intel investor, I look forward to seeing whether Kanter's predictions ultimately pan out.