Seeking Alpha
Long/short equity, contrarian, independent research, tech
Profile| Send Message| ()  

The day that Apple (AAPL) started designing its own system-on-chip ("SoC") products for its iPhone and iPad line of products, it became a competitor to Intel. Apple made a number of acquisitions and has been hiring aggressively enough in the SoC/processor design space. While previously, Apple more or less assembled off-the-shelf ARM (ARMH) and Imagination Technologies IP blocks, optimized them, and then had the chips fabricated at Samsung (OTC:SSNLF), Apple made a bold step with its latest "A6" and "A6X" line of products. It designed its own processor core, making it a bona fide competitor to Intel (INTC), Qualcomm (QCOM), and in some sense ARM itself.

Fabrication Plant Utilization Expected To Return Organically

Intel designs and builds chips for everything from phones to big-iron servers. Intel's fat margins and wild success have been due not only to its superior chip design prowess in the majority of its operating segments, but due to owning its own fabrication plants and being able to tightly integrate both (and collect the margins from both). As long as Intel can keep the fabs utilized, it collects the design margin as well as the foundry margin. The problem lately is that PC demand has slumped, so the fabs are going underutilized to clear out inventory that was built ahead of an expected Back-to-School ramp that never came.

That being said, Intel's upcoming product cycle should really ramp and should allow for a stronger than seasonal second half. This also means that fabrication plant utilization should be back to ~90%, so this isn't really a case of Intel "desperately" needing to fill excess capacity:

Interviewer: Is there enough business, enough customers to soak up the capacity of those plants? Do you believe that will be the case in 2013?

CFO Stacy Smith: Oh, yes. We expect that over the course of 2013, getting into the back half, our factories are running full, our cost structure looks great, and gross margins are going to move back into the low 60's.

Interviewer: When you say full, do you mean 98%? Do you mean 100%?

CFO Stacy Smith: Full for us is 90%-ish percent, that tends to be "full". We always want to have a little bit extra so that we can respond to upside in demand, new customers, or things like that.

It is clear that Intel is simply trying to take advantage of its manufacturing lead to add an additional revenue/profit growth stream. This, to me, means that Intel isn't exactly begging to build competitors' chips. In fact, it makes absolutely no economic/financial sense to do.

Intel Wants To Sell Atom Into Phones, Not Build Someone Else's

While building chips for strategic/non-competitor entities is a great way to add a stable revenue/profit stream, Intel ultimately wants to get paid the design margin, too, if it can. The idea is to build its own chips and then sell them to the phone vendors. A point that I have tried to make clear is that if Intel is selling $20 chips, then it makes much more margin on it than the competition, and certainly much more than if it were building a chip for someone else who then proceeds to go sell it for $20. If the business is viable for Qualcomm or MediaTek, then it can be even more so for Intel, given the right volumes.

The goal for Intel, then, is to not give its competitors access to leading edge process technology who can then use it to gain market share, but to instead build its own chips that benefit from the process technology leadership. It's a simple yet effective strategy that can certainly work in the tablets (where modems aren't a huge concern), and will work in phones if Intel's modem division can get up to speed.

So, No Apple In Intel Fabs

It therefore makes absolutely no sense to build Apple's chips on the advanced fabrication process and effectively kill the major advantage that Intel's own design teams will have over every one of the competition. Apple sells, what? 125 million iPhones and 58 million iPads? That's ~200M units @ $10/pop for Intel (Apple isn't going to be paying Intel $20/chip, since that's what the merchant vendors charge the device vendors). $2B per year if Intel scoops up every last iPhone and iPad in the world.

Is $2B/yr at 40% gross margin worth essentially giving up the ability to enable Apple's competitors (who are much more numerous worldwide) for many times the sales and ~60% gross margin? Unless Apple's going to buy Intel-designed chips from Intel outright, then no-way is Apple going to get Intel built silicon inside of its iDevices.

Source: Intel Won't Build Apple's iPhone Chips

Additional disclosure: I am short ARMH