Comms chips

How Intel Can Avoid Botching Infineon

Intel Corp. (Nasdaq: INTC) must avoid the temptation to meddle if it's to make the most of its US$1.4 billion Infineon Technologies AG (NYSE/Frankfurt: IFX) acquisition, analysts say. (See Intel to Acquire Infineon's Wireless Biz.)

So far, so good. In announcing the deal Monday, Intel said it would keep Infineon's wireless division as a standalone business. (See Intel Looks to Infineon for the Full SOC.)

There's a practical aspect to that, since the Infineon business generated $1.13 billion of revenues in its most recent full fiscal year by selling to companies that use processors other than Intel's.

But whether the Infineon group stays standalone in the long run is a crucial question.

"Intel needs to say to the new team, 'Here is an Atom CPU. Bitte, build the best smartphone product you can using this CPU. Danke,'" writes analyst Linley Gwennap of The Linley Group , in an email to Light Reading. "If Intel tries to micromanage the project, it won't work."

Intel has a long track record of failed acquisitions, especially in cases where it tries to get into new markets. Will Strauss, principal analyst with Forward Concepts Co. , counts $10 billion in acquisitions from 1999 to 2005, some of which were eventually sold off for practically nothing. (See Intel's ROI.)

The tally includes Intel's push into communications during the dotcom bubble. (See Intel Hands Off to Cortina and Marvell Takes a Bit of Intel.)

Another way to look at it: Intel's 15 largest acquisitions, excluding the recent Wind River deal, were shut down or sold on, as Gwennap noted in a blog entry yesterday.

What keeps going wrong? Some of it was bubble exuberance; Intel fell as hard for optical as any company circa 2000. (See Intel's 10-Gig Shopping Spree.)

But it's also a case of Intel imposing its thinking and methods onto the companies it buys, Strauss says. "The Intel coprorate culture kind of smothered the innovative spirit of the startups they acquired."

To avoid having the same thing happen to the Infineon business a few years from now, Intel could try stepping out of the way. The German engineering team is arguably the most valuable part of the deal, Gwennap notes, and Intel is going to have to keep from alienating them.

"To make this deal work, Intel has to respect the Infineon team's ideas and culture," Gwennap writes via email. "Infineon knows a lot more about the smartphone market than Intel does, but I am not sure Intel is willing to admit that."

Along similar lines, Intel says it would keep McAfee Inc. (NYSE: MFE) separate after the companies close their $7.7 billion merger. But that's an easier case, considering parts of McAfee's business are alien to Intel: It's hard to picture Intel telling McAfee how to sell boxed retail software at Wal-Mart Stores Inc. .

— Craig Matsumoto, West Coast Editor, Light Reading

Pete Baldwin 12/5/2012 | 4:25:08 PM
re: How Intel Can Avoid Botching Infineon

Pogo, great post. It may turn out that you've hit on the only correct answer -- stop using Atom -- in which case Intel is going to be severely tested by this deal.

This is a big one, too.  I can totally see the smartphone/tablet supplanting the laptop eventually. Intel needs an "in."

pogo 12/5/2012 | 4:25:08 PM
re: How Intel Can Avoid Botching Infineon

You know the fable of the scorpion that hitches a ride across the river on the turtle's back?  The scorpion stings the turtle half way across.  The dying turtle asks why the scorpion stung because now they would both die.  The scorpion replied, "I'm a scorpion. It's my nature to sting."

It's Intel's nature to meddle.  They might change now, but their history says otherwise.

I like Linley's advice, but even that is not enough.  Intel should give the former Infineon unit free rein and let them do whatever their market requires, even if they use a competing processor.  If Atom is the right answer, so be it.  But if it's not, they should do what is right for their market and customers.  If they don't chose Atom, Intel could learn a lot from the lesson.  If Intel wants Atom to win, they need to learn how to attract customers.  Even Intel can't buy everybody and make them switch to Atom.

Gabriel Brown 12/5/2012 | 4:25:07 PM
re: How Intel Can Avoid Botching Infineon

MeeGo running on Atom: http://carrypad.com/2010/07/13/aava-smartphone-meego-1-1-hands-on/

Not saying it's going to work, but is it that far-fetched? (I guess most people think so).

Re-posted from Top 10 Non-Android Devices to Watch 

Hanover_Fist 12/5/2012 | 4:25:06 PM
re: How Intel Can Avoid Botching Infineon

The typical Intel mindset is based on the assumption of "It doesn't exist unless we create or invent it."

Case in point...this is probably the biggest reason Intel totally doesn't "get" the high-end graphics market - their response is "we own the desktop graphics market," however nobody who plays computer games does so using the standard, off the shelf Intel embedded graphics chip - they buy third party boards.  They've been downplaying the NVIDIA and ATI graphics cards for years now and they purposely screw up their own motherboards to make sure they don't work with the ATI/AMD Graphics cards.

Now...take the Infineon purchase...within days/weeks, the mucky mucks at Intel will start convincing themselves that they actually invented Wireless and therefore have the right to "tell" the Infineon people how they're doing everything wrong...and therefore the superior minds within Intel will HAVE to start changing everything and hence destroy all that was good of the original aquisition.  The Intel marketing folks will also start telling the market place how wrong it is in their beliefs and try to strong arm people into implementing it the Intel way...vPro anyone?

It will fail miserably and Intel will be forced to sell the utterly ruined Infineon line at a deep discount (nee loss) but Intel will write that off with the utterance, "Well, they (the stupid customers) just don't get our brillance."

joset01 12/5/2012 | 4:25:06 PM
re: How Intel Can Avoid Botching Infineon

They seem to have a reasonable amount of integration work still to do with MeeGo and optimizing the battery for mobile usage if that article is on the money.

Mandrew 12/5/2012 | 4:25:04 PM
re: How Intel Can Avoid Botching Infineon

Hmm.. Interesting.

You quoted Will Strauss who thought that "Intel is getting smarter by the minute" after the vxTel acquisition.

See this http://findarticles.com/p/articles/mi_m0EKF/is_10_47/ai_71560487/

Pete Baldwin 12/5/2012 | 4:25:03 PM
re: How Intel Can Avoid Botching Infineon

Well, of course he did. I thought so too. Because Intel was doing the obvious thing: Getting into communications.

It didn't work. Strauss learned from that. Did Intel?

jepovic 12/5/2012 | 4:24:57 PM
re: How Intel Can Avoid Botching Infineon

You might be right, but what you're suggesting is that Intel would act as a pure investor. Of course, that's not Intel's business model. So you are absolutely right, it's in their nature. If Intel did not see any gain in integrating Infineon, then the deal doesn't make any sense.

Research has proven that mergers usually don't work. But there is just too much to gain for all the people involved at the top: Managers, consultants, banks and lawyers.

pogo 12/5/2012 | 4:24:57 PM
re: How Intel Can Avoid Botching Infineon


Well, it's reasonable that Intel should want its acquisitions to use Intel IP.  You are right that Intel did not buy the Infineon unit purely as a stand-alone investment.  I would certainly expect them to nudge Infineon in that direction.

On the other hand, a new processor is often a major switching cost for customers.  They have to change tools, port operating systems, work through a new batch of bugs, worry about backward compatibility, etc.  This kind of disruption is the perfect time to look for another solution.  If you have to go through all of this pain regardless, why not consider all options?  I'm not sure the Infineon baseband is sticky enough to withstand that kind of re-examination.  Wired magazine just did an article a month ago on AT&T's network problems.  In the article, they say the Infineon chip causes many of the dropped-call problems - it's full of bugs and not robust enough for the North American market in which towers are further apart than Infineon's usual European market.

With Atom, Intel needs to create a superior product and then seduce customers into using it.  Eventually, Atom will be a superior product.  However, coming from the PC monopoly/duopoly, I'm not sure Intel will ever learn the art of seduction.  They are too accustomed to dictating terms to everyone, including customers.  The rest of us realize that vendors have to grovel at the customer's feet, doing whatever the customer wants in order to get the design win.  Intel doesn't do that, so they always have trouble outside of the PC market.

Sign In