The term was coined by legendary investor Peter Lynch in his book, "One Up Wall Street," where he suggested that a business that diversifies too widely, risks destroying their original business, because management time, energy and resources are diverted from the original investment.
I just posted another comment as to what IMO Intel should be doing instead. They should "stick to the knitting". E.g. they could try to become a world class foundry. They've got the fabs, they've got thousands of great engineers. I guess that's just not as sexy as a watch that also measures your perspiration.
Intel's diversification probably has more to do with internalization of the mantra "if you're not growing, you're dying" than pressure from Wall Street.
Apple's smart growth is much better than the flailing growth that many companies attempt. E.g. Google has search so Microsoft does Bing. Apple has the iPod so Microsoft does the Zune.
New products are OK if they are clearly better than what is already out there, but in Microsoft's case they weren't. Contrast with the iPhone. It was IMO clearly better than what was already out there, if for no other reason than Apple refused to be the wireless carriers' bitch.
But now even Apple is doing share buybacks, because they're running out of smart diversification ideas. An iWatch wouldn't be big enough to provide meaningful growth. I'm sure they'll eventually do it, but the law of large numbers is no longer on Apple's side.
The example I gave of me wanting Intel to become a world class foundry is IMO smart diversification. TSM's market cap is about $93 billion. So, clearly, there's room for competition.
To me it just sounds like they hired a marketing firm to get people to pick Intel over AMD when they don't actually care what they're getting. Nothing wrong with that.
(As for mobile, Intel is getting their lunch eaten by ARM, so I can see why they might want to put more skin in the mobile game. Although, as I understand it, Intel just sells ARM manufacturers time in their last-gen fabs, thus making ARM cheaper than current-gen Intel stuff, but with not as much raw power. ARM cores make up for lack of raw power with lots of special functionality on-die, like video decoders and encoders, which your x86 Intel chip will just do in software.)
Even if they were fabbing some ARM chip, I don't think it could be one of the big players. For example Qualcomm, Nvidia, and TI use TSMC. Apple and Samsung use Samsung's fabs.
You're mostly right about it not being one of the big players. Altera is a big player in the FPGA market, but not the ARM market. The chips in question cost thousands of dollars; the ARM core is a very small portion of the chip.
http://www.nytimes.com/2014/02/16/technology/intels-sharp-ey...
I suspect that Intel is buying companies and products like this for the same reason that Microsoft bought Nokia -- they see some markets moving away from them, and rather than convincing other manufacturers and fighting competing inertia, they're going to try to make the market themselves.
That isn't to be dismissive of Intel at all -- they are a tremendous company, and their products are virtually always class leading (in capability and quality). But they were making so much money on the fat processor market they got lured into the classic fallacy of thinking that they were their own biggest competitor, and now the whole ARM ecosystem has turned into a remarkably capable foe.