I guess TI's internal discussion ran like this:
in 2015, smartphone market will be 1B+ units,
Apple/Samsung will be 500M, Huawei will be 100M, all 3 are vertically integrated with their own OS and AP, so 600M units is already out of the question.
for the remaining 400M units, at least 80% will be integrated baseband+AP for the mid-to-low-end, so TI is out.
for the last 80M units AP, TI should compete with half a dozen strong competitors including Qualcomm, Mediatek, Spreadtrum, Broadcom, Nvidia, Marvell, etc. And I just forgot Intel.Also RDA will be a strong low-end AP competitor in 2015.
Apple and Samsung don't seem to agree with Will Strauss' prediction. Do we really have a trend toward app processor/baseband integration if the top two suppliers (representing 49% of smartphones sold) don't participate. The real trend here is vertical integration.
PS- Will Strauss of Forward Concepts just reminded me that he has been predicting this since 2010. The trend is toward integrating the apps processor with the base band in smartphones, and with TI stopping development of baseband, the company has no chance. Qualcomm is already shipping Snapdragon with integrated baseband, and Nvidia plans to ship a Tegra with integrated baseband starting next year.
Personally I was surprised about this, as OMAP appears to be scoring design wins, particularly in tablets. But the more I think about it, the more this makes sense. It's a crowded market, Apple and Samsung aren't buying, and the development is expensive. Seems OMAP could have more upside in embedded.