About a year ago, Intel (INTC -1.60%) announced its next generation stand-alone cellular modem known as the XMM 7360. At the time, the modem seemed as though it had the right specifications to go toe-to-toe with Qualcomm's (QCOM -2.53%) then-upcoming MDM9x45.
Since then, however, Qualcomm flipped the script on Intel and upgraded the specifications of its MDM9x45 to support much faster download and upload speeds. Given that Qualcomm is generally regarded as the industry leader in cellular modems, it isn't a surprise it was able to "one up" a company like Intel.
What's interesting, though, is Qualcomm isn't the only company that's about to begin shipping modems with support for category 12/13 LTE upload/download speeds. Samsung (NASDAQOTH: SSNLF) just announced it would begin production of an integrated applications processor and modem with support for the same super-fast speeds that Qualcomm promises with its latest modem.
In other words, in a fashion typical for Intel's mobile group, the company has now fallen woefully behind in LTE capability relative to its competition in the world of cellular modem technology. I now believe Intel will generate very little in the way of sales with this product. Here's why.
A stand-alone modem without leadership features is a dud
In the mobile market, a cellular modem can either be integrated directly with the applications processor or it can be a stand-alone chip in a device. Generally the industry has been moving toward integrated solutions since they are usually more efficient and require less of a physical footprint on a smartphone's logic board.
The biggest customer of stand-alone applications processor solutions by far is Apple (AAPL -0.81%) since the company doesn't (yet?) have an internally designed modem that it can integrate into its A-series processors.
However, outside of Apple, the majority of phones use chips that integrate the modem with the applications processor. Most phone vendors don't sell ultrahigh-end devices and for them cost is a key consideration, which makes integrated solutions a must in the midrange and low-end smartphone markets. The single-chip solution not only reduces the number of components required and the board area, but the integration with the applications processor also leads to some power efficiency benefits.
So, who's going to buy a 7360?
The only "big" customer that comes to mind for the 7360 is Apple and given that the 7360 is behind the MDM9x45 modem from Qualcomm in terms of features as well as manufacturing technology, it seems unlikely the chip will win a spot inside of Apple's iPhone 7 next year.
For non-Apple high-end phones, the Qualcomm Snapdragon 820 should the chip-of-choice for most vendors since it should have excellent applications processor performance and best-in-class modem features (ahead of XMM 7360 on the modem side at the very least, and likely ahead of Intel's Broxton high-end applications processor as well).
For Samsung and Huawei, both of which have fairly substantial in-house applications processor and modem efforts, Intel's XMM 7360 is unattractive for their upcoming high-end phones given that both of them will have 14/16-nanometer applications processors with integrated modems (and in Samsung's case the modem is more fully featured).
This leaves very little in the way of a served addressable market for these chips. I'm going to make the prediction here now: XMM 7360, like all of Intel's other mobile-oriented chips to date -- applications processor or modem -- will generate very little in the way of revenue. I further believe that whatever revenue it does actually bring in will be at very low margins.
Obviously for low-end and midrange devices, a stand-alone modem like the XMM 7360 is a no-go.
Intel's mobile group never fails to disappoint and it would seem that, once again, letting lapped by its competitors is par for the course for this division. Maybe the company will have better luck next time?
