Comms chips

Intel Tries the Comms Market, Again

Intel Corp. (Nasdaq: INTC) has announced plans to acquire Wind River Systems Inc. for roughly $884 million, the microprocessor giant's latest strike at embedded systems, a market that includes mobile handsets and other communications gear.

The deal, announced today and expected to close during the summer, has Intel offering $11.50 per share for Wind River.

In midday trading, Wind River shares were up $3.51 (44%) at $11.51, possibly signifying that some investors expect higher bids to come in.

Intel investors were more nonchalant; that stock was up 4 cents (0.3%) at $15.98.

Wind River provides operating systems and other software for embedded systems -- a catch-all category comprising pretty much anything that has a computer in it but isn't a PC or server. Examples include cash registers, aerospace gear, medical equipment -- or switches, routers, base stations, and mobile handsets.

For all its microprocessor muscle, Intel isn't a powerhouse in embedded processors. The mobile handset market, in particular, has gone to the likes of ARM Ltd. , Qualcomm Inc. (Nasdaq: QCOM), and Texas Instruments Inc. (NYSE: TXN).

At the high-end, embedded processing has been the purview of Cavium Inc. (Nasdaq: CAVM), Freescale Semiconductor Inc. , and RMI Corp. , which is being acquired by NetLogic Microsystems Inc. (Nasdaq: NETL). (See Cavium Sprouts More Cores, RMI, Cavium Compare at the Core , and NetLogic Joins Processor Race.)

"Intel just doesn't have a very good track record with anything beyond the x86 market," the PC-processor franchise that included the Pentium, says Will Strauss, principal analyst with Forward Concepts Co. "I think they really feel they have to go beyond x86 for the next decade."

While the Wind River deal could target a wide scope of applications, it's the communications and mobility possibilities that stand out.

"With the emergence of netbooks and smartphones and so-called smartbooks -- mobile Internet devices, as Intel would call them -- the world wants to go portable, and the world wants communications," Strauss says.

But the last time Intel saw that trend, it led to disaster. "In communications, they just died," Strauss says, referring to what he counts as $10 billion in failed Intel acquisitions between 1999 and 2005. Still, Intel recognizes the need for mobility. It's put lots of money into supporting WiMax. It's introduced the Atom line of processors for mobile devices and is encouraging other companies to build systems-on-chips (SoCs) around Atoms. (See Little Laptops Could Drive Big Data Usage.)

Wind River offers software, including a Linux kernel, that could run on Atoms. Wind River also provides software that runs on digital signal processors from companies including Texas Instruments, which is another way the acquisition would give Intel more presence in the cellphone market.

— Craig Matsumoto, West Coast Editor, Light Reading

paolo.franzoi 12/5/2012 | 4:03:24 PM
re: Intel Tries the Comms Market, Again


Feel free to add to this list of Embedded things that Intel abandoned:

8085 and all its peripherals






The now Cortina assets


The hits just keep on coming.

Duh! 12/5/2012 | 4:03:23 PM
re: Intel Tries the Comms Market, Again

Considering Intel's other anti-trust problems, it's interesting that they're trying to do an acquisition that will have such a serious impact on processor vendors in a space which is actually competitive.  Particularly since the new anti-trust chief says she's serious about doing her job.  Intel lawyers must have approved the deal, but that doesn't mean that it's going to go easily.  Or that it should.

AutoDog 12/5/2012 | 4:03:23 PM
re: Intel Tries the Comms Market, Again

Wireless ASICs from Stanford Telecom

Wireless ASICs from Iospan Wireless

All optical components

Pete Baldwin 12/5/2012 | 4:03:21 PM
re: Intel Tries the Comms Market, Again

Antitrust is an interesting question here.  I asked Will Strauss about it but I already knew the answer: Embedded is a much more fragmented market than the PC, so it probably won't get the same scrutiny.

I've also gotten the impression over the years that vertical integration doesn't bother the antitrust folks at all. (There must be a counterexample out there...)

But this does present the awkward situation of Intel becoming a supplier to TI and the like.  Doesn't seem like a stable long-term situation.  Maybe different flavors of realtime Linux are the winners here?

Pete Baldwin 12/5/2012 | 4:03:21 PM
re: Intel Tries the Comms Market, Again

Thanks for the list, 7 and AutoDog.  What do you supposed Intel could do to make embedded "stick?"  (Or is their heart just really never in it?)

Duh! 12/5/2012 | 4:03:20 PM
re: Intel Tries the Comms Market, Again

Yes, but there is a new sheriff in town.  Here's what she had to say:

"On the civil front, the Antitrust Division will continue its push forward with merger and non-merger investigations. In particular, it is my hope that the Antitrust Division, drawing upon the significant expertise of my new leadership team, will have the opportunity to explore vertical theories and other new areas of civil enforcement, such as those arising in high-tech and Internet-based markets. Increasingly, Americans are relying on high-tech solutions in the home and the workplace and enjoying the fruits of innovation in those markets that have been spurred on by competition between rival firms. We thus plan to devote attention to understanding the unique competition-related issues posed by these markets. In the past, the Antitrust Division was a leader in its enforcement efforts in technology industries, and I believe we will take this mantle again. In so doing, I am cognizant that we must find the right balance to ensure that when intellectual property is at issue, competition is not thwarted through its misuse or illegal extension."  Christine Varney, remarks prepared for the US Chamber of Commerce, May 12 2009.

It will be interesting to see what happens.

paolo.franzoi 12/5/2012 | 4:03:20 PM
re: Intel Tries the Comms Market, Again


Intel goes through oscillations.  For a while, they "FOCUS" - which means dumping everything not related to PCs and making PC like things.  Then, they say "WOW - There is a whole embedded world we are not getting money from".  So, they go after the embedded market.

It seems that these things are about on 5 year cycles.  The problem the FOCUS time is trying to solve is that incrementally Intel does not command as big a set of market shares outside the PC business.  So they do the old #1/#2 or out thing. 

The "WOW Embedded" part of the cycle is to address issues in the PC growth business.  Intel is profitable but seems to be growth challenged (aka Intel's Growth is tied to Computing Growth directly).  So they go after "new markets".

I am guessing they are tied to a consulting company or a strategic planning cycle that does this.  The humorous bit is that younger folk don't realize the long history (since say 85) of Intel's abandoning various embedded markets.  So, they buy into the NXP or Optical from Intel.  As an old fogey, I just toss them out on their ear.  Unless forced to (like say when they bought Level 1), I would not base an embedded system around parts from Intel.



schlettie 12/5/2012 | 4:03:19 PM
re: Intel Tries the Comms Market, Again


PCIe Advanced Switching

Infiniband (not embedded, specifically)

sigint 12/5/2012 | 4:03:18 PM
re: Intel Tries the Comms Market, Again

It seems that these things are about on 5 year cycles.  The problem the FOCUS time is trying to solve is that incrementally Intel does not command as big a set of market shares outside the PC business.  So they do the old #1/#2 or out thing.


I have been witness to this from close quarters. For a peculiar mix of reasons, Intel seems particularly inept at integrating acquisitions and product lines. For starters, any non-processor work is completely secondary at Intel. The best engineers and executives from acquired companies are often forcibly *redeployed* to the processor world, pretty much setting the original group for failure. And then, deadwood from the processor world floats in, ostensibly to *integrate* the acquisition. They then lead the group to ruin, and conclude that Intel should have never got into this business anyway and stayed focussed on processors!

Another redeployment, laid off engineers, and one more good company and product line gone forever.

Repeat ad nauseum, at 5 year intervals, as 7 pointed out.

It's a great fab company, and they should build chips for whoever needs them. That's what they do best.

HomerJ 12/5/2012 | 4:03:08 PM
re: Intel Tries the Comms Market, Again Doesn't Windriver rely on partnering with the embedded processor guys like IBM, Fresscale and TI? I'm no rocket surgeon but I'm thinking they will be somewhat hesitant to "partner" with Intel. Why exactly are they doing this? The embedded guys have been burned by Intel waaaay too many times. This is good news for the Linux guys and the likes of QNX.
Sign In