Intel recently announced in its 10-K filing that it will move away from its "tick-tock" chip production cycle and toward a three-step development process instead. For decades, Intel's chipmaking business has relied on alternating "ticks" and "tocks" every year.
Continue Reading Below
During a "tick" launch, Intel introduces a smaller microarchitecture with a new manufacturing process. During a "tock" launch, the microprocessors are upgraded, but the size remains the same. From now on, Intel will replace that 24-month cycle with a 30-month "process, architecture, optimization" model instead.
Inside an Intel manufacturing plant. Source: Intel.
The original tick-tock strategy formed the foundation of Moore's Law, a concept named after Intel co-founder Gordon Moore, which claims that the number of transistors within the same area of silicon will double every two years. Following that "law" kept Intel at the top of the PC and server chip markets, while serving as a plot device for countless sci-fi stories about malevolent AIs.
But in recent years, it's become much harder to shrink down chips and squeeze in more transistors. During a conference call last year, CEO Brian Krzanich admitted that the tick-tock period had expanded closer to two and a half years rather than two. Therefore, Intel's official expansion of the tick-tock cycle into three steps wasn't that surprising. But will doing so dull Intel's competitive edge against rival chipmakers?
Understanding the new 30-month cycle
Intel's "process, architecture, and optimization" model basically adds an additional "optimization" stage to the end of the tick-tock cycle. Simply put, it lets Intel stay on each "tick" -- as defined by the size of the die, in nanometers -- for longer than before.
Continue Reading Below
Intel's newest generation chips are 14nm ones. Intel ushered in that generation with the Broadwell design in September 2014, which it upgraded in August 2015 with Skylake. Intel will then "optimize" the 14nm design with Kaby Lake near the end of 2016. Intel originally intended for Skylake to be directly succeeded by the 10nm Cannonlake chips this year, but the company has since delayed that "tick" upgrade to thesecond half of 2017.
Intel insists that this expanded launch cycle doesn't mean that Moore's Law has ended. Critics, however, have long claimed that Moore's Law is merelya business and marketing strategy instead of an actual scientific "law". For example, the original Moore's Law from 1965 called for the number of transistors to double every year, before it was stretched to a "two-year cadence" in 1975. The official expansion of that "law" to 30 months indicates that the 1975 version of Moore's Law needs to be revised again.
A single wafer consists of multiple squares, or "dies." Source: Intel.
What does this mean for Intel's business?
The good news for Intel investors is that the chipmaker still creates industry standard, "best in breed" chips for PCs and servers. AMD , its only meaningful competitor in the x86 arena, shot itself in the foot when its current generation of Bulldozer CPUs fell behind Intel's in terms of single-threaded performance.
AMD's best hope for striking back is launching its new generation of Zen chips by the end of this year. However, the Zen will be AMD's first 14nm CPU, and it will arrive more than two years after Intel's Broadwell. The Zen might compete effectively against Intel's Kaby Lake processors, but it's highly doubtful that it will launch 10nm processors before Intel.
Last summer, IBM apparently leapfrogged over Intel with the creation of a 7nm chip. However, that process used a silicon-germanium manufacturing process instead of the pure silicon process which Intel uses. Therefore, the chip was an interesting showcase piece that generated a lot of buzz, but it's still far too expensive to be fully commercialized. Moreover, IBM's server chip designs use the Power architecture, which has been marginalized by Intel's 99% share of the data center market.
Intel investors have other things to worry about
The "end" of the tick-tock cycle and Moore's Law might reduce the speed at which Intel shrinks its die, but it's hard to see AMD catching up and denting its share of the PC and data center markets anytime soon.Instead, investors should focus on the near-term threats facing the company -- sluggish PC sales and its failure to enter the mobile chip market. Both factors will likely weigh down its Client Computing business, which posted an 8% sales decline last year.
Introducing new chip architectures every two to three years is still a solid long-term plan for Intel, but it's unclear if those new chips will convince stubborn PC users to consistently upgrade their aging, "good enough" machines.
The article As Moore's Law Ends, What's Next For Intel Corporation? originally appeared on Fool.com.
Leo Sun has no position in any stocks mentioned. The Motley Fool recommends Intel. Try any of our Foolish newsletter services free for 30 days. We Fools may not all hold the same opinions, but we all believe that considering a diverse range of insights makes us better investors. The Motley Fool has a disclosure policy.
Copyright 1995 - 2016 The Motley Fool, LLC. All rights reserved. The Motley Fool has a disclosure policy.