Although Intel (INTC 1.77%) is sure to share additional details of its current strategy to attack the market for mobile processors in about a month, company CEO Brian Krzanich offered some interesting insight into its strategy on its Oct. 13 earnings call.
Let's take a look at what Krzanich had to say.
Modems for the general smartphone market; partnerships for "specific products"
On the call, Krzanich said the company's longer-term strategy to attack the "general phone market" is to bring out stand-alone cellular modems at an annual cadence. These stand-alone modems would presumably be paired either with Intel's own applications processors (ranging from smartphones to laptops) or with applications processors designed by others.
Then, per Krzanich, for "specific products" Intel will apparently pursue a strategy in which it co-designs chips with partners, such as Spreadtrum and Rockchip. In this case, the products are not just individual modems but are integrated system-on-chip designs that integrate both the applications processor and the cellular baseband processor.
The problem with the stand-alone modem strategy
The issue that I see with the company's strategy of going after the "general phone market" with just stand-alone modems is that the industry as a whole is moving away from stand-alone modems and toward integrated applications processors and basebands.
The majority of the smartphones shipped today, from low-cost phones to premium flagships, use processors with integrated cellular baseband processors. Only a select few, albeit fairly high volume, devices such as the iPhone and some variants of Galaxy flagships, use stand-alone applications processors and modems.
In essence, for Intel's stand-alone modem strategy to work, it will have to aggressively fight for -- and win -- spots inside of these few flagships. And, even then, longer term I expect Apple and Samsung will transition to integrated solutions.
So, if Intel can land a spot inside of Apple's iPhone at some point down the line, then that could be a nice revenue stream, but it would be a fairly low-quality one as it may eventually disappear.
What about this "partnership" thing?
What Intel is doing with its aforementioned partnerships is a bit more nuanced. The way that it has been explained previously is that Intel provides companies like Spreadtrum and Rockchip with fundamental system-on-chip building blocks and even a baseline design to work from.
From there, Intel works with its partners to build system-on-chip products targeted at particular markets and price points. Intel and its partners would then be able to sell these chips.
Intel hasn't disclosed to investors what the financial arrangements are between Intel and its partners in this case, but I suspect that over the long term, the financial relationship will be similar to that of a foundry/fabless customer.
Will this strategy actually work?
At the end of the day, Intel executives have said that the goal in mobile is to make money. Given that Intel's mobile group -- before it was folded in with the company's PC client group -- lost $4.2 billion in 2014, the company certainly has a long way to go to get there.
If Intel can get very large volumes of mobile processors into its factories (i.e., gaining large market share by way of its partners) and collect essentially what amounts to the kind of margin dollars that a third-party foundry would get, then I could see a path to profitability for the company's mobile group.
This wouldn't be a particularly glamorous way to make money in the mobile market, but the company might be far better off pursuing this strategy rather than trying to go after the market alone and ultimately failing as it has over the last four years or so.
