Intel’s Manufacturing Lead Explained

The calls from Wall Street keep coming with basically the same set of questions: “Does Intel really have a 2-3 year process lead? Can Intel lead the foundry segment? Can Intel Lead the Mobile SoC Segment?” The feeling amongst the buy and sell side investment people is that unless Intel can lead a market they will not stay in it (Intel TV). Given that, my answers are no, no, and no, absolutely.

The calls I get from Wall Street are generally a direct result of articles published on Motley Fool and Seeking Alpha. They pretty much let anyone post on these sites and Seeking Alpha even lets you post under pseudonyms. From what I have read there is very little if any actual semiconductor experience behind these articles yet they talk semiconductor with authority. They all have disclaimers at the bottom saying what stocks they own but that gets lost when people start quoting and cutting and pasting from the articles.

Disinformation of course is a competitive weapon and something Intel is very good at. Most of these authors are spoon fed PR stuff, they Google the rest to support what’s in the spoon. Add in the personal bias of owning the stock and you get a serious amount of disinformation which is why Wall Street keeps calling.

But of course I have a bias too. I do not trade stocks but having grown up in the fabless semiconductor ecosystem I’m very biased for it. Semiwiki even published a book on it “Fabless: The Transformation of the Semiconductor industry.” That is why I get the calls, so they can hear both sides of an issue, which can lead to a much more informed investment decision.

The real game changer in the semiconductor industry is not Intel but Samsung. Samsung is going after Intel to be the world’s largest semiconductor company and I believe it will happen, absolutely. Just look at their financials and aggressive marketing tactics. This company is out to win at all costs.

Samsung’s recent announcements with GLOBALFOUNDRIES and STMicroelectronics provide additional pieces of that roadmap. Samsung is now second sourcing a 28nm FD-SOI solution which is above and beyond what TSMC offers and Samsung is now driving 14nm process development rather than following IBM.

The advantage of working in Silicon Valley for 30+ years, meeting people face-to-face, and developing mutually beneficial relationships is access and honesty. After talking to both TSMC and Samsung about FinFETs I can tell you with a great amount of certainty that designs have taped-out, yield is better than expected, and risk production will start in Q4 2014. TSMC and Samsung will both ship close to 1M FinFET wafers in 2015.

Intel has struggled with 14nm of course as reported accurately on SemiWiki last year. Let’s not forget that Intel CEO BrianK held up a laptop at last year’s Intel Developers Forum and crowed, “14nm is up and running!” Today Intel claims 14nm will be “up and running” in 2H 2014 in the shape of Broadwell (PC) and CherryTrail (SoC). If that is in fact true, Intel has a six month manufacturing lead at 14nm. I wonder what BrianK will hold up at this year’s IDF?

It’s a shame Intel does not disclose wafer shipments so we can do an apples to apples comparison here. How about it Intel? Show us a little FinFET transparency? How many 14nm wafers are you going to ship next year? Of course you would have to break out microprocessor and SoC wafers because as we all know SoCs are a much harder to produce.

Bottom Line: Intel’s purposed manufacturing lead is a paper tiger, absolutely.

Also Read: Intel’s SoC Challenge!

More Articles by Daniel Nenni…..

lang: en_US