Intel (INTC +8.78%) boldly proclaimed that it would be selling "at least" 40 million tablet chips throughout calendar year 2014. While getting there will require the company to forgo any real revenue or profit on the chips (thanks to some hefty subsidies), this can be viewed as a relationship-building exercise with the major OEMs. The expectation, of course, is that when the products have the "right" cost structure come 2015, Intel can unlock substantial profitability here. With that in mind, the big question is whether Intel can do the same thing with smartphones.
It's not quite as straightforward
Intel has proven that it can hang with the best of the ARM crowd when it comes to tablet performance and power consumption. Further, since tablets don't usually ship with cellular technology, Intel's lack of an integrated cellular baseband in its products isn't much of a barrier to winning designs in this space (particularly since Intel is shipping discrete data LTE products today). However, the smartphone space is quite a different animal.
In this space, of course, the applications processor is important. Qualcomm (QCOM 1.17%), the leader in this market, clearly puts a great deal into research and development, but the modem is arguably more important. During 2014, Intel will have a lineup of fairly competitive applications processors with "Merrifield" and its quad-core cousin, "Moorefield." Its XMM 7260 discrete modem looks good, but how will it hold up to what market-leader Qualcomm will have?
It is likely that, thanks to the low-power 22-nanometer process and exceptional Silvermont CPU design, Intel's apps processors for phones will be lower power than what Qualcomm is able to do with its 28-nanometer Snapdragon parts (the highest end of which have already garnered a reputation for throttling back performance in phones due to heat). However, next year, Qualcomm will still have an advantage on integration (Qualcomm's high-end Snapdragons integrate Wi-Fi, Bluetooth, and cellular baseband). While this isn't likely to be a deal breaker for Intel, it's going to make it tough to convince OEMs to switch to Intel.
Could Intel pull the same trick twice?
The big question is whether Intel is going to try to pull the same trick with smartphones as it is doing with tablets. While Intel's smartphone platforms will be designed for smartphones, the bill of materials issue that plagued Bay Trail's assault into the low end won't be here for Merrifield/Moorefield. However, converting designs that are currently in the clutches of Qualcomm will be difficult, unless Intel is willing to price incredibly aggressively.
Further, the bigger concern is the modem technology. Intel's modem products do not include CDMA support, so many of the phone vendors that are used to Qualcomm's parts across their entire lineups would have to do two potential SKUs: one with an Intel apps processor and Intel modem, and one with either an Intel apps processor and Qualcomm modem or a Qualcomm integrated apps processor and modem. It is likely that OEMs would choose the integrated Qualcomm solution.
Finally, Qualcomm is aggressively taking content share in phones by pushing its own connectivity combo chip solutions as well as its own power management ICs and audio CODECs. In short, Qualcomm can offer "the complete platform" to handset OEMs, while Intel still isn't quite there yet. Indeed, Intel has yet to ship its own low-power connectivity combo chips (although word on the street is that they'll be coming at some point next year).
Foolish bottom line
While the tablet market is very much like the PC market in that integration isn't as important and providing the "complete" solution isn't absolutely critical, the smartphone market is an entirely different beast. With that in mind, it is likely that 2014 will be a year in which Intel wins a few "important" smartphone designs from the likes of Google/Motorola, LG, and HTC. But, the company will still have a tough time displacing Qualcomm in the flagship designs until 2015.
