Is Intel Selling Technology It Doesn't Have? - Seeking Alpha

Rethink Technology business briefs for September 20, 2017

Intel holds another Technology and Manufacturing Day in China

Source: Intel

Intel (NASDAQ:INTC) already held a Technology and Manufacturing Day event last March, so its second event in China has received scant attention in the media. After writing about the March event, I was curious to see what, if anything, was different about this new Technology and Manufacturing Day in China.

There was quite a bit of new material, and I'll go over the highlights. As in the past, Intel presented valuable technical information, and then marred it with flimsy assertions and marketing spin. With each of these presentations, going back to the February Investor Meeting, the desperation to assert process leadership seems to ratchet up a notch. As the claims become more strident and self-serving, they also become less credible.

The emphasis was also slightly different. TMD1 was about persuading US investors and analysts that Intel still has a competitive edge. TMD2 was about persuading the Chinese fabless semiconductor industry that the company is a viable foundry alternative to Taiwan Semiconductor Manufacturing Company (NYSE:TSM) and Samsung (OTC:SSNLF). So, Intel led with a complement to the China semiconductor market:

Intel's Stacy Smith, president of Manufacturing, Operations, and Sales, made the point that the foundry TAM is split between leading edge (28 nm and below) and older process nodes

He also stated 70% of the leading edge market is controlled by one foundry, and that Intel was specifically targeting that foundry. That foundry is of course, TSMC, the first and largest “pure play” foundry name. Given that the audience was primarily Chinese, I'm really not sure how well that went over. It conjured up images of heavy-handed Westerners attempting to muscle their way into the Chinese market.

Most of the day's presentations were devoted to trying to prove how superior Intel's 10 nm technology is compared to competing 10 nm processes. Intel once again claims, with reasonable justification, that its 10 nm process offers nearly twice the transistor density of competing processes.

But note that as things stand right now, TSMC's and Samsung's 10 nm processes offer greater transistor density, by Intel's own elaborate definition, than Intel's 14 nm process.

Reports of further delays in 10 nm

Therein lies the crux of the problem for the company. Intel's case for its process superiority would have been so much more compelling if it actually had shipped 10 nm products. Samsung has been making 10 nm Exynos and Snapdragon parts since Q1 of this year. TSMC followed in Q2 with the Apple (NASDAQ:AAPL) A10X for iPad Pro and the A11 for the new iPhones.

Intel is trying to make the case that it is maintaining the Moore's Law pace of doubling transistor density every two years, despite having been on 14 nm for three years. The company argues that it's doing this by taking greater steps between process nodes and maintaining a 3-year process lead.

The claim of still being ahead by 3 years will almost certainly be proved wrong in 2018. In terms of actual shipping products, TSMC and Samsung are ahead of Intel right now. Even accepting Intel's assertion that competitors' processes are one generation behind, both TSMC and Samsung are expected to have their 7 nm processes in production by late 2018. Even with the Intel “correction”, they will be roughly equivalent to its 10 nm process, assuming that Intel ever gets 10 nm in production.

During the presentation, the company tantalized the audience with the promise of production start in 2017, and displayed some 10 nm wafers.

Yet, there is already a report from DigiTimes that Cannon Lake has been delayed until the end of 2018. Cannon Lake was expected to be the first wave of 10 nm processors, since they are relatively small devices intended for low-power mobile applications.

One indicator of desperation in technology companies is that they start selling technology they don't have, or don't quite have. Even now, Intel doesn't quite seem to have its 10 nm process ready for production.

Intel's “collaboration” with Waymo

Another indicator of desperation in technology companies is a tendency to announce “collaborations” that are less than they seem. CEO Brian Krzanich's blog post (labeled an Editorial) titled “Waymo and Intel Collaborate on Self-Driving Car Technology” is one such announcement.

The article is full of interesting, though not very useful, prognostications about the future of self-driving cars. However, the gist of the collaboration is that Waymo uses Intel processors and other technologies such as FPGAs and cellular modems in its current fleet of test vehicles.

Waymo's Twitter page offers a slightly different take on the purported collaboration with statements such as:

Waymo's compute platform inside our minivan feature Intel-tech to help make real-time driving decisions.

Our compute is designed in-house, letting us work directly with partners like Intel.

If Waymo is designing the computer in-house, then what is Intel's contribution? Krzanich makes it seem that Intel is leading the charge in autonomous vehicles, when it's really just a component supplier, and not even the only one.

Those of us who have been following the technology since the DARPA Grand Challenge days know that early developers often opted for various PCs as the host for their software. They were cheap, powerful, and components were readily available. It sounds as if Waymo has taken that a step further, designing its own logic boards as it does for its data center servers.

But it's still basically a PC in the trunk solution. That works for demonstrations and test vehicles. It may even work for ride sharing services. But it's unlikely to work for automakers looking for compact, low-power processing platforms.

While it's easy to postulate an Intel/Waymo dominance of the self-driving future, that's unlikely to happen with the current technology they deploy. Automakers have already been casting their votes for what they think is the hardware solution of the future, Nvidia's (NASDAQ:NVDA) Drive PX Xavier.

Nvidia is part of the Rethink Technology Portfolio and is a recommended Buy.

Disclosure: I am/we are long NVDA, TSM, AAPL.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.

http://ift.tt/2fcr6lA