Last week, an article on CIO.com caught my attention: ARM will skip modems for now, despite being a huge force in mobile. I can only applaud them for this, as all too many companies have a hard time focusing on the things they do well--in ARMs case, creating an licensing designs for mass market, special purpose processor designs.
The problem with ARM also including modem chip designs is not that they won't be any good at it: it's directly in line with the current solutions they build (including Bluetooth designs), and will undoubtably be a huge revenue driver. In stead, there is a mismatch in business models: whereas ARM is very good a creating a small number of very high quality designs that are then licensed, diversifying into the modem market would lead them into a maze of different network technologies, regulations (more so than with Bluetooth hardware) and customers with very differing wishes, requiring close collaboration with each customer for each batch.
All in all, this decision is not about technology, but about alignment. ARM is the company that paused to wonder not whether it could, but whether it should.