Another Good Intel Idea Bites The Dust

Covad Communications, Clearwire, SpectraWatt – what have they all got in common? All were backed by Intel. And all failed.

But they all seemed a good idea at the time.

Covad was one of the ‘CLECs’ – competititive local exchange carriers whose existence was enabled by the US 1996 Telecommunications Act.


Setting up in the wired telecommunications business in opposition to the traditional carriers seemed like a good idea at the time. Founded in 1996, Covad was sold to a private equity company in 2008.


Clearwire set out to provide wireless data services in the US using Wimax. Intel put $1.6 billion into the venture in 2009 and wrote off 90% of it.  Wimax is basically seen as a dead dog in the US as LTE takes over.


Spectrawatt spun off from Intel in 2008 with $50 million of Intel VC money to manufacture solar cells. It has closed its factory and filed for bankruptcy.


What is it about Intel? Does inhabiting a quasi-monopoly cocoon make it difficult to operate in freely competitive markets?



  1. Despite the long list of Intel failures, one should see that Intel has succeeded in one, major, huge way:
    Monopolizing the industry through a secret, worldwide scheme that regulators blew the top off of last year.
    Several billions in bribes to Dell.
    Hundreds of millions (according to subpoenaed documents summarized by the NY AG) paid to HP to keep AMD under 5%.
    Now why in the hay would a monopolist bribe a manufacturer to “MOSTLY starve” AMD?
    Well, that’s Intel’s biggest problem. Killing AMD gets Intel in trouble. Allowing AMD to make enough money to pay their R&D debts and hire new employees gets Intel in trouble, too.
    So… Intel shall probably continue to secretly bribe and threaten manufacturers to sell just a few AMD products here and there.
    Monopolistic Corruption Perfected. No need to innovate. The cocoon is all Intel needs.

  2. Don’t forget that Covad failed because Mr. Powell’s FCC ransacked the 1996 Telecommunications Act and handed monopoly status in wired broadband to the ILECs. The GOP loves big business lobbyists.
    As for WiMAX Sprint is the only big carrier interested and they have bungled their 2G and 3G efforts so you could say that WiMAX is not well represented. All the same, it seems that LTE is technically better on several points though more expensive to roll out, so it’s the almost certain winner in developed countries. The lower entry cost of WiMAX does seem to be getting some interest from large organizations for private WWANs though.
    As for Intel, it’s the usual problem with them: hubris, which could be related to the “cocoon”.

  3. A fine distinction, Mr C, the differentiation eludes me.

  4. Mr M, I do apologise I should have said “what an XXXX creeper”

  5. The plain unvarnished truth, I can assure you, Mr C.

  6. Oh Mr M what a XXXX creeper?

  7. With some that might be true, Mr C, but not with Mike

  8. “Yes, Mike but, with the greatest respect”
    Come on Mr M we all know that it translates to “with no respect”?

  9. Yes, Mike but, with the greatest respect, I would point out that this list is compiled by Intel which isn’t, as we know from the anti-trust cases, unfailingly committed to telling the plain, unvarnished truth at all times and, in the list of ‘successful exits’ , is Clearwire on which Intel wrote off 90% of its $1billion investment. So I approach this list with more than one grain of salt.

  10. See
    I suspect most VCs would love to have this number of IPOs which is where they make real money. And a couple of the private sales were still good deals.

  11. Torben,
    Torben, we are talking about acquisitions. USB and PCIe, for example, were homegrown inventions.

  12. Intel had the Xscale processor platform 10years ago (it bought the strongARM platform from DEC in 1998) and were pushing it into handheld applications among other things.
    Several PDAs and Blackberrys used the Xscale at the core of their products. However Intel decided that the x86 market was their core competence and sold the Xscale to Marvell in 2006.
    I remember thinking back in 2001 that the Xscale would make for a fantastic mobile computing platform. The problem was that none of Intel’s customers managed to capture the consumer’s mind with their devices and so the revenue generated by the platform compared to the R&D and marketing costs didn’t stack up well compared to the x86.
    Had they waited just a little longer they now would have a great platform for the mobile market.
    It was a case of too soon in the market, too soon out of it. Now they’re trying to re-invent the wheel when they could “simply” make use of their ARM license… You’ve got to wonder at times.

  13. In defence of Intel, they have also backed successes. USB, for example.
    A company as big as Intel will back a lot of technologies and will expect only a few of these to succeed. But one big success can pay for a dozen failures, so it all adds up in the end.
    What Intel has really failed to do, however, is to make a processor design to replace x86. They tried with iAPX 432, which was a dismal failure, and they tried with Itanium, which didn’t fare much better (The nickname “Itanic” says it all).
    Of course, Intel didn’t have a big incentive to find a replacement when x86 dominated both the PC and server markets. But ARM is encroaching in both and the portable market gets increasingly important. Intel _is_ trying to get into the portable market, but apart from the increasingly irrelevant netbooks, they haven’t had much success.

  14. I’ve never worked for Intel, but the horror stories that I’ve heard, from good engineers that get caught up in Intel communications acquisitions, defy all logic. Such fundamentally stupid actions, that one truly has to wonder, why Intel made the acquisitions in the first place. The most common refrain is, WT* were they thinking?
    Whenever I hear about new Intel comms product acquisitions, I’m reminded of a sign I read in a Texas High school football stadium.
    “Welcome to the Boneyard”

Leave a Reply

Your email address will not be published. Required fields are marked *