Does Intel have a strategy for IoT?

You don’t have to look past the $4.2 billion that Intel’s mobile business lost last year to be reminded how difficult playing catch up can be. The fact that Intel lost the smart phone and tablet market early on to ARM has reverberated with the company for years.
And as we approach smartphone saturation in the developed world in the next few years, all semiconductor companies are looking for where the next major market will come from. Intel is no exception.
Whether that next major market will turn out to be the Internet of Things is one of the biggest questions right now. Surely, connectivity is proliferating and the 50 billion device figure by 2020 is likely to come true. We are entering a massive age of connectivity. And to that end, business models designed around monetizing that connectivity are apace. Every company wants to be part of the hardware solution for those 50 billion devices.
The problem for Intel is that many of these embedded solutions are similar to mobile in that they are power sensitive and space sensitive. Intel’s focus on building power efficient processors has greatly improved due to the lessons of mobile but the company still has a long history of building fast, high margin, high power processors for the desktop and server market.
To some extent, one of the issues that could actually impact Intel is how much processing in IoT winds up being done at the edge device to eliminate latency problems (think issues related to autonomous driving) versus how much gets done at the back end server. Broadly IoT favors Intel’s efforts in the data center, which saw 18 percent growth in 2014 for the company, as more connectivity requires greater cloud services. But if greater processing power is required at the end device that could create a greater opportunity for chips with more robust processing capabilities and help Intel.
Still, Intel isn’t waiting to find out how things play out. At CES this year, it unveiled Curie, a chip that integrates a processor, Bluetooth LE, sensors and an engine for determining sporting activity. Eying the wearables market, the integrated communications/processor/sensors system in a very small form factor will be attractive to hardware developers, particularly on the higher end. It is slated to ship before the end of the year.
Intel is pegging single digit growth for the company this year to its data center and internet of things business. Intel CFO Stacy Smith told CNET that he has “no idea” how large a market exists for Curie.
He said that he thought “Curie is a great proof point of how that can evolve. That’s a fully capable computer that can connect that was in the size of the button.” And on IoT, he was similarly cautious, noting that “[IoT] is a very chaotic segment. There’s lots of innovation going on. If any one tells you they know what’s going to be the winner three years from now, they’re making it up because nobody knows. We want to be there early. We want to be there early with great technology.”
To some extent, Smith is right to be cautious. It’s early in IoT. But in other respects it’s increasingly clear that lower end markets like those for microcontrollers are growing because more and more devices are requiring small, low power processors. What’s less clear is whether Intel processors will find a home in the market for higher end connected devices like some wearables that may need a bit more processing power. This will also get increasingly competitive if multicore microcontrollers evolve and are able to meet the demands of device makers.
IoT will, no doubt, drive growth for Intel, at the very least at the data center/server level. But if the company finds a way into more embedded devices, it would be a welcome surprise for investors who watched as the company lost out on the last most promising semiconductor market—smartphones.