dont count on WiFi or BLE or any consumer-based communcation protocol: they are not designed for cellular-graded reliability or power IoT consumption.
I agree with the cellular approach to be the right choice for IoT, but using 3G is an overshoot. we've developed a IoT modem in real silicon with only 10mm^2 which can form the celluar-like link with 50,000 nodes with very low power consumption. 300mm^2 with 3G modem and multi-chip package? definitely not for the fire alarm or sensor.
Ain't that small is relative. This is a stepping stone. Clearly IoT modems need to get smaller, but the comm link question will mostly drive that. If you need 3G (do you?) then ok. But if you can get by with BTLE or something with a smaller silicon footprint, then even better.
IoT is a very big market. It can create the same revolution as PPC market once created. Intel if puts in right time and money in IOt chip development they cn retain theirn victorious position for decades to come.
@Jessica Lipsky - "IOT" is so generic. Cellular communications is one of several methods that are already being used. I do not see any real alternative to this technology for some use cases. As a side note, take a look at the M2M market. There are many similarities to IOT. You'll notice the market growth. Intel can't ignore it.
I think most people realize that thinking of IoT in terms of one technology is problematic. WiFi, BLE and Cellular are complementary and there are applications where one is better than the others. I know because at Argenox Technologies we work with customers using all of these and that is when you realize that to build a real application you need a technology that will fit. As an example, one customer of ours uses Cellular because deploying Zigbee requires a density of devices that may not be achievable in this fixed infrastructure mode. It is more costly, but the reliability is something they can't afford to mess with. In another case, a product we designed with WiFi makes sense becase it is in a home or another location where WiFi is obvious. Zigbee may also work, but there are other considerations such as throughput.
If Intel has customers that would buy this chip, that's great. if they designed this for IoT without a specific market/application then they may never recoup their investment, unless it is done for marketing or as a proof of concept.
IoT is something that will come by itself when things are connected. Trying to connect everything when it isn't needed is wasting money. There is little value in connecting a dinner plate to the cloud (a bit of exxageration but I have seen worse suggestions). There is value in connecting a garage door. Companies will do well focusing on real applications where there is benefit (lowering energy consumption, enabling remote monitoring, etc). Then companies such as Intel, TI, Freescale, ST, Qualcomm, Broadcom, etc can provide devices that help make the connection cheaper and better.
It's a perfect technology pick by Intel, Cellular Communication will be in demand for IoT as most part of the world is today covered by cellular technology. Intel will be surely working on the missing links in this modem will very soon come with the entire solution range for IoT Modems. There are many other uses of this chip will be there, it can be a part of traditional basic mobile handset, GSM Modem for Internet access etc.