Intel delays work on 10nm Cannon Lake chips, plans third round of 14nm CPUs

By Brad Bourque July 15, 2015

three generations intel hd graphics tested version 1454824832 intelchip — Shutterstock

Back in 1965, Intel co-founder Gordon Moore predicted famously that transistor density would double every year for ten years. In 1975, he revisited that prediction and revised it to state that transistor density had slowed to doubling every two years. Now, it looks like Intel is doubling back once again, with or without Moore, and extending that timeline, at least temporarily, to every two and a half years.

In an investor call regarding the second quarter of 2015, CEO Brian Krzanich announced that the 10nm chips planned for early 2017 will be moving back to late 2017, and there will be one more round of 14nm chips after Skylake, simply called “Lake,” according to ZDNet. This is likely the Kaby Lake architecture leaked late last month.

It’s no secret that technological transitions to smaller and denser platforms have slowed in the last few years. Intel managed to make the move to 14nm with the Broadwell chips in the second half of 2014, while AMD still struggles to go any smaller than the 28nm chips it introduced years ago. That’s not entirely AMD’s fault, as it not longer owns its production facilities. Still, it’s clear all companies that design chips are facing a common problem.

Though almost no details are known about the Lake chips at this time, the core is likely to be responsible for only a small number of processors, and possibly the fastest of that socket, if Intel doesn’t change it again. Other than that, all we have to go off of is Krzanich’s comment that Lake will be built on the same design principles as Skylake, but will add in “key performance enhancements.”

The good news is that Skylake-based CPUs are still very much on track for the second half of 2015. We’ve already seen a few of the motherboards that are ready for the Skylake launch, as well as a couple of enthusiast CPUs built on that architecture. We don’t have an exact release date yet, but when we do we’ll be able to more fully understand the Intel roadmap for the next two years.

Editors' Recommendations

Topics

Former Digital Trends Contributor

Brad Bourque is a native Portlander, devout nerd, and craft beer enthusiast. He studied creative writing at Willamette…

Computing

Intel Alder Lake-S leak: New hybrid 10nm architecture features 16 cores

Benchmarks for Intel's unannounced 12th-Gen Alder Lake-S desktop processors have leaked, suggesting that the company may be progressing nicely with its silicon development. Alder Lake-S will be Intel's first desktop processor to switch to a 10nm node, so gamers can likely expect some performance uplifts, but the chip will also bring more architectural changes as well, like the adoption of the big.LITTLE core design from mobile.

Even though the Geekbench benchmark results spotted by Twitter user @TUM_APISAK revealed that the 12th-gen processor tested contained 16 cores and 24 threads, the core configuration is different from Intel's current desktop processors or even rival AMD's Ryzen 5000 CPUs. Though unconfirmed, rather than equipping Alder Lake with 16 total high-performance cores, Intel is said to be configuring this chipset with eight Golden Cove cores with 16 threads and eight Gracemont cores with eight threads, Wccftech reported.
Golden Cove and Gracemont in action

Computing

Intel CFO: “10nm isn’t going to be the best node that Intel has ever had”

intel ice lake wont rid spectre insecure chip

Intel is looking to jump past prior woes with its 10nm processor architecture to a smaller, more efficient 7nm or 5nm design in an effort to compete against rival AMD's latest 7nm CPU and GPU offerings.

"Look, [10nm] just isn’t going to be the best node that Intel has ever had,” Intel CFO George Davis admitted to analysts at Morgan Stanley's Technology, Media, and Telecom conference, according to a report on German publication Computer Base. "It’s going to be less productive than 14[nm], less productive than 22[nm]."

Computing

Intel’s 10th-gen Comet Lake chips could be more expensive than ninth-gen

Intel Announces The Xeon 5100 Microprocessor For Servers

The first hint of pricing for Intel's upcoming 10th generation of Comet Lake CPUs suggests that the new generation could be more expensive than the last across the board. Listings for a number of entry-level and midrange 10th-gen CPUs have appeared on Czech and Slovakian digital stores, and the prices are between 2 and 15 euros higher than they were at the launch of Intel's ninth-generation Coffee Lake CPUs.

Intel has been forced into an unfamiliar position in recent years. Following more than a half-decade of performance dominance in almost all sectors of the market, which let it almost arbitrarily set its prices as it wished, it found itself under threat. AMD's Ryzen processors first nipped at its heels, and then overtook it in instructions per clock and core count, leading to Intel slashing prices on its HEDT chips, and cutting them lightly on its more mainstream parts. For the first time in a long time, it was Intel that had to appeal on value, where it couldn't on performance.