Recently I there was news that Texas Instruments was laying of 1700 people and re-focusing their OMAP (ARM-based platforms) on embedded and away from mobile. Back in the 1990s when I was at VLSI and working closely with our communication division, TI was the giant in the industry. They were Nokia's biggest supplier and we were completely unsuccessful at gettting any of their business. It is obviously not as dramatic as a Kodak going backrupt (and doing it like a slow train wreck since it is not as if even Kodak hadn't seen digital coming, they pioneered in the area) but we have the same creative destruction in the semiconductor industry. Of course part of the reason may be that Nokia itself, which at the start of the year was still the worlds large cell-phone manufacturer has fallen so dramatically. In smartphones it is hanging on by its fingernails in position #10 having been #3 last quarter. I fully expect it to be sold although it would have been worth a lot more 6 months ago before they had their own round after round of layoffs. Interestingly, Motorola (which Google bought) has completely fallen out of the top 10 smartphone vendors, and it will be interesting to see how Google responds (since everyone largely assumed Google bought Motorola mostly for patents).

Intel announced that Paul Ortellini will be stepping down in May as CEO a couple of years ahead of expected. The general word on the street is that the board decided that they need a CEO who is going to make Intel a success in mobile, otherwise Intel will be suffering its own slide into obsolescence as gradually the PC fades. It won't go away, of course, but future growth is in tablets and smartphones. In fact some time next year there will be more smartphone in the world than other types of computer. Well, not all the embedded microcontrollers, of course. But servers, desktops, laptops, tablets, iPads and all. That's right, even counting iPads and Kindles on the computer side of the balance. Almost all of them are ARM-based, all most all of them contain chips manufactured not by Intel but by TSMC. I've said since the beginning that I thought Intel's Atom strategy would fail, in the sense that they would not ship a lot of chips.

The challenge for everyone is that smartphones are dominated by Samsung and Apple. They are not the only players but they make almost all the money. And they build their own application processors. Qualcomm is the leader in the cellular chips. But that means that the merchant part of the market, where at TI or an Intel builds chips and sells them to cell-phone manufacturers is not that big a business. Only the weakest competitors are in the market. It is not a small business but you generally want to be selling to the market leaders.