Intel clarifies 14nm products and 10nm in 2H'17 (earnings conference webcast 7/15/15)

[pxc]

From zdnet and the conference call (live webcast ended a few minutes ago and download should be available soon):

UPDATE: During the shareholders call on Wednesday, Krzanich provided a little more color on Intel's chip roadmap with an explainer rooted in Moore's Law:

Just last quarter, we sell celebrated the 50th anniversary of Moore's Law. In 1965, when Gordon's paper was first published, he predicted a doubling of transistor density every year for at least the next 10 years. His prediction proved to be right and in fact, in 1975 looking ahead to the next 10 years, he updated his estimate to a doubling every 24 months. These transitions are a natural part of the history of Moore's Law and are a byproduct of the technical challenges of shrinking transistors while ensuring they could be manufactured in high volume.

As node transitions lengthened, we adapted our approach. This strategy created better products for our customers and a competitive advantage for Intel. It also disproved the death of Moore's Law predictions many times over. The last two technology transitions have signaled that our cadence today is closer to two and-a-half years than two.
Thus, following Skylake, Intel is preparing to trot out a third 14-nanometer product with the simpler moniker "Lake." Lake is said to be built on the same foundation of Skylake but "with key performance enhancements."​

A 10-nanometer model with the code name Cannon Lake is promised to follow during the second half of 2017.

In response to some push back from analysts and investors curious about the delay for the 10-nanometer edition, clarified that Intel hopes to bring the time frame back to two years eventually.

"Remember on all of these technologies, each one has kind of its own recipe of complexity and difficulty. 14 to 10, same thing that happened when 22 to 14," Krzanich remarked.

So Intel confirms there's going to be a slight delay for 10nm, but it's still coming in 2017 (~2.5 years between 14nm and 10nm).

 

[obiwansotti]

And we already something like this between 22 and 14 with the Ivy Bridge, Haswell, Haswell "Devi's Canyon" cadence, followed by the non-launch of the broadwell parts.

 

[pxc]

And we already something like this between 22 and 14 with the Ivy Bridge, Haswell, Haswell "Devi's Canyon" cadence, followed by the non-launch of the broadwell parts.

Broadwell was launched late last year with Core M (Broadwell-Y). More interesting mobile parts were launched a few months later and OEM systems with Broadwell-H started appearing a couple of weeks ago.

It's strange that boxed processors are so delayed, but that's hardly a "non-launch".

 

[obiwansotti]

Broadwell was launched late last year with Core M (Broadwell-Y). More interesting mobile parts were launched a few months later and OEM systems with Broadwell-H started appearing a couple of weeks ago.

It's strange that boxed processors are so delayed, but that's hardly a "non-launch".

That's exactly what I meant by a non-launch though, it just sort of quietly started appearing and replacing things.

 

[pxc]

That's exactly what I meant by a non-launch though, it just sort of quietly started appearing and replacing things.

I think that's really a problem created by (almost always inaccurate) rumor sites. Intel announced launch windows, and announced delays when necessary. Launching Broadwell-Y before the end of the year was to meet that promised delivery window (Q4'14), and delays for other mobile parts were announced months earlier. Desktop parts launched in the time frame Intel promised. Like I said, boxed processor availability is strangely late, but that's a tiny portion of Broadwell's launch.

My speculation: today's Intel financial results and webcast reference almost $600 million in "IoT Platform and other" sales, which is crazy. That includes server infrastructure. I think Intel is using 14nm capacity to service that segment as a priority (more profitable 1S Xeons) over desktop parts.