Following up on its earlier release of 8th generation Core processors for mobile PCs, Intel this week extended the product line to the desktop.

But as I noted in my previous write-up about the new mobile processors, Intel is confusing the marketing on this new generation of chips. So I’ve done something I rarely do, and consulted with an expert so that I could better understand—and communicate—what’s really happening.

“With the 8th generation Core processors, Intel decided to break it up within each product segment,” explained Ryan Shrout, an analyst at Shrout Research. “On the mobile side, the 8th generation processors were codenamed Kaby Lake Refresh … On the desktop, the chips were codenamed Coffee Lake, but they are built on the same architecture. They could have—maybe should have—called these chips Kaby Lake Refresh too.”

Put another way, the 8th generation Core processors on mobile and desktop respectively provide similar benefits over their predecessors.

On mobile, Intel went from a dual-core design, with Hyper-Threading to a quad-core design with Hyper-Threading, but using the same 15-watt power envelope. “That’s actually a pretty interesting shift,” he told me. “It’s not quite double the performance, but it is in the 70-to-80 percent range, and it happens with no decrease in single-threaded performance.”

That last bit is important, both technically and for your understanding of these chips as a potential purchaser of a new PC: The 8th generation Intel Core processors offer little in the way of performance improvements over their predecessors when it comes to typical productivity tasks. The real improvements are in multi-threaded activities. Things like 3D work, video editing, Lightroom, Photoshop, and so on.

On the desktop, the newly-released Coffee Lake chips—also part of Intel’s 8th generation family of products—offers similar improvements. In this case, Intel has taken the chips from four cores to six, necessitating a bigger die size. “Note that this is the same micro-architecture as Kaby Lake,” Shrout told me. “Nothing has changed.”

So why did Intel use a new name, Coffee Lake, for these desktop chips? Why not just use Kaby Lake Refresh here as well?

“Intel went with the Coffee Lake branding because the processor requires a new chipset,” he told me. “You can’t put an older Kaby Lake chip in the new motherboards. And you can’t likewise put a new Coffee Lake chip in your existing motherboard.” Apparently, a pin change related to power delivery is to blame.

There are some interesting additional notes about the mobile and desktop 8th generation processors, too.

You may have noticed that Dell, HP, and other PC makers have already announced new portable PCs based on this updated design. And that, less publicly, PC makers are also updating existing products—as HP is doing with its 15-inch Spectre x360—to utilize the new quad-core Kaby Lake Refresh parts. But the availability of these systems is limited, and it seems like only the biggest PC makers are on-board so far.

“It seems like Intel is holding back these parts for a handful of their biggest partners only,” Shrout noted. “They are aimed at flagship PCs only, and we’re not going to see these chips in a wide range of systems anytime soon.”

The reason? Cost. As it turns out, cranking out quad-core Kaby Lake Refresh processors costs Intel significantly more money than making dual-core Kaby Lake processors. But Intel isn’t charging much more for these chips.

This raises another obvious question: Why on earth would Intel not charge more—maybe much more—for these more powerful, arguably mobile workstation-class, chips?

Because of AMD. With this other microprocessor maker on the verge of releasing mobile versions of its well-regarded Ryzen chips this fall, Intel needed to do something to blunt the impact of its suddenly resurgent rival. And Kaby Lake Refresh is how Intel can do that.

“The Coffee Lake parts are also a bit of a reaction to AMD,” Shrout said, “and while some people don’t like that term, that’s how competition works. It’s a back and forth. But Intel has a real engineering advantage too. So it can react very aggressively.”

In any event, I’ve seen a lot of hyperbole around the new 8th generation Intel Core chips aimed at the desktop, for example, Gizmodo declared that they are “as fast as hell.” That’s actually true for certain workloads. But it’s also true that the major advances you get with this generation are only a minor (10 percent-ish) speed bump in typical usage. And that it requires a motherboard change if you’re upgrading. All of which sort of undercuts the click-bait headline, I know.

But Intel does appear to be providing a lot of value at a time in which competitors old (AMD) and new (Qualcomm) are turning up the heat. I may disagree with the marketing, but Intel’s technology is clearly top-notch.

“Numbers and branding don’t matter,” Shrout explained. “The performance of the technology does. How high it clocks, how dense, how much power leakage. Intel is able to hit 4.7 GHz on the desktop with 6 cores and yet remain very power efficient. They just have a huge lead.”

Tagged with Intel, Intel x86