Earlier, I took at look at what went wrong with Intel's (INTC +23.60%) smartphone plans during 2014. But as investors, it's important to understand but not dwell on the past. As Intel looks to take on mobile juggernaut Qualcomm (QCOM +11.12%), it's worth taking a look at what Intel has in the pipeline for 2015 and how that will stack up with what Qualcomm has in store.
The apps processor roadmap
At Intel's 2013 Investor Meeting, management presented the following plan for 2015:
Source: Intel
You'll notice that at the top, Intel plans to release a product codenamed Broxton for the high end of both smartphones and tablets.
At the low end, Intel is introducing a product called SoFIA ("Smart or Feature Phone with Intel Architecture"). The details about this product are a bit more nebulous, but we do know that the 3G version will be a dual-core Silvermont core design and that the LTE version will sport four cores and a feature-reduced variant of the XMM 7260 (based on Intel statements at the 2013 Investor Meeting).
Both of these product families will be important in driving Intel's smartphone strategy, with SoFIA likely to be the more important one in the near-to-medium term.
What's the strategy?
The strategy with Broxton appears to be to design a no holds barred high-end smartphone and tablet applications processor. This should serve Intel well as it pushes into the Android and Windows tablet markets, and could help Intel win a few sockets at the likes of LG, HTC, and perhaps a few of the lesser known brands attempting to aim at the high end. The opportunity for Broxton in high-end phones probably isn't giant -- and it is dependent on Intel executing on the modem/communications side of things -- but long term, Intel is going to want to be a credible player here.
The more interesting play here is with SoFIA. In 2015, SoFIA will be built on TSMC's 28-nanometer manufacturing technology. This is likely due to the fact that the modem IP that Infineon Wireless -- the team that Intel acquired to build its modems -- had already been working on these designs on TSMC's process technology.
The good news is that this will allow Intel to be able to compete with a fully integrated part, but the bad news is that this strips Intel of its manufacturing lead until 2016. But the more important point is that the low end of the smartphone market -- where SoFIA and its follow-ons are targeted to -- isn't dominated by Samsung and Apple. The market is more diverse, and Intel can (potentially) sell to a broad range of customers.
Will it work?
Only time will tell if this strategy will work, but if Intel can actually gain traction with the 28-nanometer SoFIA parts, then once it rolls out follow-on low-cost designs on 14-nanometer in Q1 2016, Intel should have a fairly sizable performance/power/cost edge over the rest of the competition. Assuming that Intel integrates competitive connectivity and cellular IP, Intel should be able to gain reasonable traction in the low-end of the smartphone market at good margins.
Where's Qualcomm in all of this?
Intel doesn't exist in a vacuum and its results in phones depends pretty heavily on the competitive landscape. Qualcomm's product lineup for 2015 seems to be the following:
- Snapdragon 410-low cost, 28-nanometer LP product based on the ARM (ARMH +0.00%) Cortex A53. Integrated MDM9x25 modem and 1-stream 802.11ac Wi-Fi. (Q3 2014 device availability)
- Snapdragon 610/615 -- mid-range, 28-nanometer LP product based on 4-8 ARM Cortex A53 cores. Upgraded graphics over Snapdragon 410. Same modem/Wi-Fi as Snapdragon 410. (Q4 2014 device availability)
- Snapdragon 808 -- high end 20-nanometer part with 2 ARM Cortex A57 cores and 4 ARM Cortex A53 in big.LITTLE configuration. Adreno 418 graphics. Integrated MDM9x35 LTE-Advanced modem (no Wi-Fi). (1H 2015 device availability)
- Snapdragon 810 -- high end 20-nanometer part with 4 ARM Cortex A57 cores, 4 ARM Cortex A53 cores in big.LITTLE configuration. Adreno 430 graphics. Integrated MDM9x35 LTE-Advanced modem. (1H 2015 device availability)
The SoFIA LTE part, depending on its graphics capabilities (this is a big unknown), could either be very competitive with the Snapdragon 610/615 or the Snapdragon 410. Intel doesn't really have an answer for the Snapdragon 808 and 810 during the first half of 2015 since Broxton is a second half of 2015 event.
The bottom line is this: Intel seems as though it will be competitive with the lower end of Qualcomm's product stack in terms of performance and timing. The high end of the smartphone market is a bit trickier to call as it is unclear when Qualcomm will refresh the 808/810 products.
Foolish bottom line
The good news is that Intel should finally have credible, timely solutions for the low end of the smartphone market -- Intel's biggest opportunity in phones given current market dynamics. It will need to address this market with products built at external chip manufacturers, but if it can use 2015 to build credibility with the handset vendors, then this paves the way nicely for the higher margin 2016 parts built in Intel's manufacturing plants.
