If there's one thing that I like about Intel (NASDAQ:INTC) (and believe me, I like plenty), it's that the firm is not afraid to spend its cash to bolster its technology portfolio. While the major headline "expense" is capital expenditures, and this is closely followed by R&D to the tune of $10B+, it is interesting to see some of the smaller acquisitions that the company makes. Earlier we saw Intel gobble up engineers and assets from ZiiLABS, presumably to beef up its low power graphics portfolio, and then of course we had the Fulcrum, Cray (NASDAQ:CRAY) fabric, and QLogic (NASDAQ:QLGC) fabric acquisitions. Today, Intel picked up ST-Ericsson's GPS division/assets, and it speaks loudly and clearly to what Intel is aiming at going forward.
Intel - It Stands For Integrated Electronics
The real point of Moore's law is that on a regular cadence, semiconductor companies can pack more transistors in a given area than in previous generations and have things stay economical. What you're seeing now in the semiconductor space is a frenzied attempt from players such as Taiwan Semiconductor (NYSE:TSM) and Samsung (OTC:SSNLF) to try to "catch up" to Intel on transistor technology. While with the "16nm" and "14nm" processes from these two firms will offer minimal scaling benefits over their unreleased 20nm processes (so, think 2015/2016 before any real volume on these hybrid processes), Intel will go into high volume production of its own 14nm process that brings the traditional scaling benefits from the prior 22nm generation.
Make no mistake, this is a gigantic lead that will only get wider. This means that at the same die size, Intel will be able to pack on significantly larger numbers of transistors, which means that each chip can contain more "value." While the transistors at the high end are typically spent making these chips faster, putting in more cores, larger caches, and so on, the real value of a process lead in smartphones/tablets parts and other highly integrated chips is that Intel can integrate more things and consume a larger part of the bill of materials. This means higher selling prices for each chip.
Broadcom (BRCM) is a company that has seen rather nice success in the smartphone space, selling WiFi/Bluetooth "combo" chips. Qualcomm (QCOM) too has seen success in the mobile apps processor space as it has integrated the modem among other components onto its SoCs. Intel, too, plans to release an integrated baseband + apps processor in 2014. But this is just the beginning.
See, the value that having the world's "best" transistors isn't even necessarily for performance (although this is still a key consideration), but for integration. If you can integrate more onto your chip than your competitors while at the same time keeping your die sizes smaller, then this is an instant competitive advantage. Not only can you actually command an ASP premium to your competitors, but your production costs are lower. This is precisely how Intel appears to be using its process technology lead in the low power space, and I think that it will become even more evident to the general investment public as Intel's first 22nm parts hit the market later this year.
The acquisition of GPS technology from ST-Ericsson means that the firm is looking to integrate GPS integrated circuits onto its SoCs fairly soon. In fact, did anybody else catch that Intel's Atom is already used in quite a few in-vehicle infotainment systems and that this is a market that the firm continues to emphasize in its future plans? Of course, GPS makes sense for handsets and tablets, too.
We're Already Seeing It In Graphics
We've already seen this before. Intel now uses GPU technology to differentiate and modulate pricing. A higher end CPU coupled with beefier graphics commands a substantial premium to a higher end CPU paired with less substantial graphics. This is a way to consume more of the bill of materials of a given system, and in return actually get paid for it. When people see that they can spend an extra $40 to get a much better gaming experience and a better processor to boot, then that becomes a way for Intel to upsell to higher end parts.
This is why you are now seeing Intel so heavily invest in graphics - it's a point of differentiation and a way to make more money given the process/node size lead.
This is the key to the Intel mobile margin puzzle. With smaller transistors, Intel can integrate more things while keeping the area the same or even smaller than its nearest competitors, all while commanding a higher price not simply due to faster CPU cores, but due to the fact that the chip consumes more of the bill of materials. Intel is a mashup of "Integrated Electronics" and in the mobile space going forward, it is very much going to be an integration race. Intel, I believe, will be a very strong player going forward, and it won't be too long before the first "Bay Trail" parts are with us, finally showing us what the company can do.
Disclosure: I am long INTC, QCOM, QLGC. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.