I know I promised that my next 2016 prediction would be Apple’s big acquisition, and I will publish that prediction soon as my #10, but right now I just have to say what a perilous position Intel is in. The company truly risks becoming irrelevant, which is an odd thing to say about a huge, rich outfit that would appear from the outside to pretty much dominate its industry -- an industry the company created. Intel won’t go away, I just think there is a very good chance it’ll no longer matter.
We’re approaching the end of the closed, proprietary, single source technology era. ARM processors are freely licensed, more open, and much more cost competitive than similar products from Intel or AMD. If you need 10 million chips for your next product do you buy them from Intel? Or do you get a license from ARM and hire a foundry to make them for you?
The same can be said of operating systems. Do you go buy 10 million licenses from Microsoft, Apple, IBM, or..? Or do you go get a blanket license for Android from Google?
The interesting questions that will determine the future are:
Will Intel start making ARM chips? It has done it before: remember StrongARM? If Intel doesn’t re-embrace ARM for at least some of its line it will be a much smaller company in a few years.
Microsoft’s CEO seems to be quite smart and a good visionary. I am more optimistic about Microsoft’s future now than I’ve been in years. Will Microsoft start making Android products and applications? Porting Office to Android/ARM will be a better strategic decision than when it ported Office to the Mac.
Back to Intel, the company made a lot of news recently by laying-out a new corporate strategy based on data centers, Internet of Things, and memory -- explicitly de-emphasizing both personal computers (in decline) and mobile (where they haven’t had much success). I think this is smart but unless Intel follows it up with better tech in the very areas ARM has come to dominate the strategy won’t work.
Just to take one example, there’s a huge opportunity in data centers, which is to say building clouds. Most commercial clouds are based on free, very cheap, and/or open source technology. They use a low cost hypervisor. The disk storage is not as fast, secure, or reliable as it needs to be. We’re going to have some technology burps along the way as these deficiencies both become known and are taken advantage of by the bad guys. When that happens we will start to question our needs -- question the cloud.
Intel can fix these problems and move on to even greater heights, but I don’t think it knows how.