Intel today is launching what it calls the first of its 8th generation Core microprocessors. But this initial release is really just core-doubled “Kaby Lake” (7th generation) processors. True 8th generation chips—“Coffee Lake”—won’t arrive until next year. And the oft-delayed move to a more efficient 10nm manufacturing process—“Cannonlake”—is further off still.
Confused? Welcome to the club. But the reasons behind this marketing disaster were triggered by two clearly-understood factors: Intel’s slowing pace and the rise of more efficient and mobile ARM chipsets.
As you may recall, Intel finally acknowledged that it was no longer able to maintain its “tick-tock” development schedule—in which major chipset releases (“ticks”) were interspersed with minor updates (“tocks”)—in early 2016. It’s new and slower-moving methodology is called “Process-Architecture-Optimization,” and it can be thought of as “tick-tock-tock” or even “tick-tock-tock-tock.”
Skylake, the disastrously buggy 6th generation Core chipset that Intel first released in 2015, was a “tock” release, and it utilized the firm’s 14nm manufacturing process. It was to be followed up by Cannonlake, Intel’s first 10nm Core chipset.
Unable to deliver on its 10nm vision, Intel instead has been forced to iterate its 14nm chips while still maintaining the annual release cadence its PC maker partners expect. So the firm launched Kaby Lake, its 7th generation Core chips, in summer 2016, offering modest performance and battery life improvements but much better reliability than its predecessor. I referred to Kaby Lake as Intel’s apology for Skylake, as Skylake bugginess was partially responsible for the reliability issues that users of Microsoft’s Surface Book and Surface Pro 4 experienced. (As we now know, Microsoft was partially to blame as well.)
Intel has further muddied the waters, however, but spreading out the release of its newer chipsets over wider periods of time. So while some Kaby Lake chips did ship in 2016, the more powerful versions, with quad-core designs aimed at desktop PCs, workstations, and high-end laptops, didn’t ship until the first half of 2017. And by mid-year, Intel was expanding the range yet again with new Xeon-like Core-X series chips that would offer up to 18 processor cores and “extreme” performance (and pricing).
And now they’re muddying the waters even further: This week’s launch of so-called 8th generation Core processors does not include true Cannonlake designs. Instead, the first 8th generation Core chips are just tweaked Kaby Lake designs or, as Intel calls them, a “Kaby Lake refresh.” (Kind of a demi-tock, I guess.) Because Intel.
On the good news front, this marketing silliness shouldn’t obscure the fact that these chips are still interesting. They are the first Core U-series processors—which are aimed at Ultrabooks and other ultra-portable PCs—to feature quad-core designs (with 8 threads), so they deliver a performance boost of up to 40 percent over the previous designs. Intel says they are shipping now, and that about 80 new PCs featuring these chips will be available over the holidays.
And since I complained about the naming, I’ll also point out that Intel, for some reason, will refer to upcoming Coffee Lake and Cannon Lake chips as 8th generation Core processors. Frankly, I expect that to change. By the time Intel does make the switch to 10nm, I suspect it will want to trumpet the newness a bit more.