Later this month, Intel is going to ship its XMM 7160 discrete LTE solution for mobile devices. With this new design, Intel will support 15 LTE bands (there are 30-40 bands, and growing), which is a good step in addressing the LTE band fragmentation that is currently causing difficulties in terms of network interoperability and worldwide compatibility for LTE devices. Unless carriers decide to use a common set of LTE bands, it will be very difficult, if not impossible (in the short term), to completely solve the LTE “fragmentation” problem, but handling 15 bands is a whole lot better than a handful, and it could allow device makers to improve coverage for a single device SKU (stock keeping unit, or “variant”). This is particularly difficult to do without increasing the board size, since additional frequencies normally requires more on-board components.
Additionally, Intel believes that its solution will end up consuming anywhere between 20% to 30% less power than the competition. Since the LTE sub-system (or radio in general) is often among the top 3 sources of power drain, any progress is highly desired in that area. Intel also points out that its solution requires a bit less space than other options OEMs may have at their disposal. XMM 7160 was announced earlier this year, and at Computex, one OEMs has already announced that it would be integrated in an upcoming product.
Modems typically don’t get the hype that System on a Chip (aka SoCs) do, but they can be more important, business-wise, than raw processing power and fancy graphics. At the moment, Qualcomm is getting a tremendous amount of business because it has great LTE products that are proven and pre-qualified with most carriers worldwide. Although the rest of their SoC is excellent as well, the modem is a crucial part of their overall success. And it completely makes sense: no matter the processing power, connectivity trumps everything else in a mobile world.
That is exactly why NVIDIA, MediaTek, Broadcom and others are bringing up their LTE solutions, whether it is discrete or integrated. Intel knows it and is executing on LTE. Intel’s customers (OEMs, carriers…) want more choice (because that would lower prices) so they are reportedly quite supportive of Intel’s efforts. While major US carriers have not announced the XMM 7160 qualifications on their network, Intel has said that it is working with all of them, and it’s fair to assume that the product will be qualified soon.
Intel knows that success will take time, and that a leadership position has to be “earned” over time. But the company made it clear that it wants to leverage the fusion of talents between Intel and Infineon (acquired by Intel in 2010) to build wireless systems that will be highly competitive and ultimately ship in volumes. When that time comes, Intel is ready to scale, big.