On Intel's (INTC 2.10%) second-quarter earnings call, CEO Brian Krzanich updated investors on the status of the company's 10-nanometer manufacturing technology. There had been much speculation in the press that Intel planned to delay its 10-nanometer ramp. Those suspicions were further confirmed when news broke that Intel planned to launch a third 14-nanometer PC processor family in 2016.
During the call, Krzanich didn't elaborate much on the issues that the company was facing with its 10-nanometer technology, other than to suggest that these issues are "similar" to what it faced in ramping 14-nanometer chips into production.
What does this mean for Intel's product line?
On the call, Krzanich said Intel planned to introduce a follow-on to its Skylake processor known as Kaby Lake in the second half of 2016. Following that, he said, will be a 10-nanometer product family known as Cannonlake in the second half of 2017.
The good news is that Intel's difficulties in ramping up its 10-nanometer technology won't leave system vendors offering "stale" Skylake-based systems in the second half of 2016; they'll be able to put together more compelling systems with Kaby Lake.
Additionally, by putting out an improved design, those Kaby Lake systems will offer more performance and potentially more features than Skylake-based systems. In a PC market that is struggling, Intel really needs to do everything in its power to make PCs more compelling so that people actually feel the urge to upgrade.
The bad news is that Intel won't be able to deliver the relatively large efficiency gains that traditionally come with a migration to new manufacturing technology until 2017.
So, about that chip manufacturing lead ...
Intel has long prided itself on its chip manufacturing lead over the rest of the industry. It's fair to say that over the past several generations, Intel has had a solid lead over the competition in terms of adopting new transistor materials and structures and in migrating to new manufacturing technologies.
With Samsung and Taiwan Semiconductor (TSM +2.73%) claiming that they will go into mass production on their respective 10-nanometer technologies by the end of 2016, it seems that the two foundries have closed the gap with Intel.
During Intel's earnings call, Krzanich addressed this point:
We believe that even with this 2017, our lead in Moore's Law will not change dramatically. We believe that we'll continue to lead with roughly the same leadership position that we have today.
We base that on, one, what really counts when I talk about 2017 that's not samples, that's not small-volume, that's converting over to Cannonlake and producing a large percentage of our CPUs in volume in the second half of 2017.
Krzanich also went on to say that it won't just be Intel's ability to deliver 10-nanometer product in high volumes.
There's this definitional difference; this will be now our third generation of FinFET, it will have several other transistor enhancements, and we believe if you take a look at the scaling it will be quite strong relative to the normal scaling parameters that occur with a Moore's Law transition.
Competitor TSMC has said that its 10-nanometer technology will provide about 2.1 times logic density scaling over its 16-nanometer FinFET Plus technology. TSMC also claims that its 10-nanometer technology will deliver a 19% performance increase at constant power or a 38% decrease in power consumption at the same performance relative to 16-nanometer FinFET Plus.
It will be interesting to see what kind of performance, power, and area gains Intel is able to deliver with its own 10-nanometer technology when it finally reaches the hands of consumers in the second half of 2017.
