Intel CEO reaches for the TV-off button to focus on more important issues

Intel CEO reaches for the TV-off button to focus on more important issues

Summary: Intel's lack of a high volume mobile chip could lead to unexpected alliances...

SHARE:
TOPICS: Intel
2

Intel's [$INTC] plans to offer a streaming TV service by the end of this year is on the ropes and its prospects have greatly diminished under new CEO Brian Krzanich.

Peter Kafka reporting on AllThingsD:Intel Wants Helps from Amazon or Samsung to Launch OnCue Web TV - Peter Kafka - Media - AllThingsD

People familiar with Intel’s plans said the company has been talking to both Amazon and Samsung in the hope of keeping the service afloat.

Foremski's Take: The streaming TV service was announced in early 2013 and is a project backed by former CEO Paul Otellini, who left the chipmaker in May.

The Intel TV venture has about 300 staff and there is a small pilot project being tested by several thousand Intel employees.

But this TV project never made any strategic sense especially since Intel has much larger challenges on its plate related to its main business of chip manufacture.

When Mr. Krzanich became CEO in May, his first interviews signaled that the TV project would not be a priority. He has now told the team that they must find partners or it will be shutdown. It's highly unlikely that Intel will get the media partners it needs to keep it going.  

Building a foundry business…

Intel's current priorities are how to build a foundry business making custom chips for clients. This is a tough business for Intel because it's always built only its own chip designs, except for a brief period when it built ARM based chips. It has tried to build a foundry business once before and abandoned it.

Another challenge facing Intel is its move to advanced chip fabs. Its chip manufacturing expertise is the world's best, and it is way ahead of rivals by a year or more.

However, it is missing something tremendously important that is absolutely needed if it is to keep ahead of rivals with the best, most advanced chip fabs. It needs a high volume, cutting-edge mobile chip to make so that it can quickly ramp up its future fabs.

Intel is not a microprocessor company…

The key to understanding Intel is understanding that it is a chip manufacturer first — it just happens to make microprocessors because they are highly profitable.

In order for it to build advanced chip fabs it needs to manufacture a chip that it can use to calibrate and fine-tune more than 300 separate manufacturing steps. Without that, its production yields will be low, and it'll take much longer to ramp up production of other chips.

With the latest fabs costing more than $9 billion, it risks massive losses if it can't put that all of that capital to work as quickly as possible. 

Why Intel needs mobile chips…

Because of the smaller markets for desktop PCs and laptops, and its failure to dominate in mobile chip markets, Intel lacks a high volume cutting edge chip to help it ramp its advanced fabs.

If it had, for example, the manufacturing contract for Apple's latest A7 iPhone chip, it would be an ideal chip for driving its advanced chip lines. However, Intel's rival Samsung has that contract. 

Former Intel CEO Paul Otellini recently said that one of his regrets was not taking the iPhone seriously when Intel had a chance to make chips for Apple. He didn't think there would be much demand and so he quoted Apple a price higher than it wanted to pay, and chip foundry TSMC won the contract. 

The iPhone became the world's biggest selling consumer digital device. Apple's revenues from its iPhone business over the past four quarters were $88.4 billion. Intel's 2012 revenues were $52.3 billion. 

Maybe the new CEO can woo Qualcomm, the top fabless chip company, as a foundry customer? It has exactly the high volume, cutting edge mobile chip designs that Intel needs.

But overcoming Intel's incredibly strong culture of competitiveness with Qualcomm, and to using the British ARM microprocessor design that's the basis of Qualcomm's designs —  would be massive cultural challenges.

The last thing Intel's CEO needs right now is figuring out how to offer a commercial streaming TV service.

Topic: Intel

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

2 comments
Log in or register to join the discussion
  • Agree on the tv scrapping, but not the rest

    I think there's plenty enough in intels current baytrail, airmont, and forward evolutions to work out their production processes. They need to get the roi on these just for the non-mobile targeting chips. Mobile is on top of that. And really it's the forward that counts. Intels already succeeding very well with high yields on 10nm on 450 wafers. This alone puts them well ahead of the field and they didn't need A7 or any other mobile to accomplish it. And from early work right now there's no concern that there will be anything preventing the same for 7nm. Producing mobile chips at high volume would be great for increasing the roi beyond projections but not required for the investment to be recouped. That they can do it for much less per chip than any other arm foundary and crank out higer perf, lower power chips than arms should get them share going forward. You can already see their mobile design wins picking up. When you're investing 10s of billions you're in a marathon, not a sprint. That the mobile devices arms are targeting are so disposable also bodes very well for them.
    Johnny Vegas
  • Out to lunch ......

    Tom, re Qualcomm: Were you on some other planet when Otellini said he would never "enable our competitors". He meant it! and Krzanich means it too.

    Intel has some perfectly good high volume products to debug its 22nm and 14nm fabs (Haswell, Bay Trail and Broadwell).

    Revenue-wise, their success depends on Windows 8.1 takeup but this looks promising:

    http://techreport.com/news/25446/300-ppi-sharp-tablet-has-bay-trail-inside-waterproofing-outside

    As for "High yields at 10nm on 450nm wafers" ... Not too sure 'bout that Johnny: you must be talking to people in Oregon that I don't know.
    P0l0nium