You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Ouch! Mentioned at the tail of another thread, but I thought I'd elevate the discussion. This puts an end to my speculation that Intel would eventually team up with Microsoft to release a Windows 10 Mobile phone. The lack of anything from them at MWC and the departure of Aicha Evans are now a lot clearer.
And do you really think Intel is going to have better luck in the merchant modem market? Intel really is behaving like Google with all of their moonshot programs. The difference being Google is in a leadership position and more honest about their failures.
When I worked at Intel they had a core value of NIH (not invented here) which made it extremely hard to get access to commercial tools and IP that would have made us more competitive because it was "the Intel way".
It really was a good laugh to see that Contra Revenue was the new "Intel way"!
Intel problems are laid bare by mobile mistakes: Time to market (slow); margin requirement (high); innovation (low); x86 ecosystem health (declining). That's the downside. The upside: All of these are 100% under INTC control. Second, INTC doesn't need mobile; it needs speed, lower margin requirements, innovation and a growing x86 ecosystem.
Speed: Separate manufacturing and design into separate businesses. Intel Design would be able to use outside foundries to lower costs and hit ship dates. And Intel Manufacturing would be a better partner for foundry customers.
Margin requirements: Focus margin requirements on new designs. Allow margins to fall as designs age.
Innovation: Intel has the best people, but they are not producing nearly as many patents as they should be. This is partly cultural--"brag sheets" should become patent applications.
x86: Everything else in tech has evolved over the years but not x86. ARM arch is only slightly better in this regard. I think Nvidia invented a better mousetrap that no one is using (ARM compatible arch); Intel should put aside NIH and license it.
Intel isn't the first company to retreat from the smartphone market. It happened to Nvidia and they had to write down their Icera modem assets. I can see a similiar thing happening for Intel's modem as it no longer serves any strategic purpose.