Having 35 years of experience in the electronics business during the "Inventor Generation" has provided me with some interesting insights from people who went on to be very influential movers and shakers in the business.
One late night discussion many, many years ago involved the idea of measuring companies by their ability and willingness to do the difficult things, and to do those difficult things first.
Xilinx (XLNX), for example would always produce the largest FPGA (Field Programmable Gate Array) in a new, proprietary, series of parts first. The rationale was that if they could successfully produce, with good yield, the largest part in the series, backfilling with the smaller parts later would be a piece of cake.
Another example would be with a manufacturer of commodity devices. I would generally try to determine whether they could service the largest and, therefore, the lowest priced piece of business that I knew of in their market. The rationale here is that if they could match the lowest price on the highest volume business, they could also handle any smaller level of business at a probable higher price. If not, they were doomed to lose the big pieces of business to the more efficient suppliers in the market.
What does this have to do with Intel (INTC)?
When Intel announces a new X86 architecture or a process shrink, the first parts available are the high end i7 devices, filled in later with the i5 and i3 devices. Much like the Xilinx example, Intel makes sure that the most complex devices work well, which makes manufacturing the less complex devices a "walk on the beach."
In the market for high performance computing chips, Intel has recently moved into production on the Xeon Phi chip. The chip is capable of "well over a teraflop." The Phi chip has 62 cores and is over a square inch in chip area. The high performance computing market is not large, but Intel needs to be involved in this computing "high ground" as a challenge to the company as a whole.
Intel, last week introduced the Lexington SoC (Z2420) with the XMM 6265 3G+ baseband modem. This platform is described as a "value solution" for the very large, price conscious, emerging markets. This multi-chip solution is about 100 square millimeters of silicon run on 32nm fully depreciated Intel production lines. The Intel cost on these chips is about $6 combined with a resale of $12 to 15 in the huge volume markets. The part has all of the functions expected for an entry level smartphone, plus it has an FM radio (poor man's iTunes). My iPhone doesn't have an FM radio! The new Intel chip can transmit content to a TV wirelessly, my iPhone can't do that either. It has dual SIM (subscriber identification), so it can be used in a two service provider mode (no contract). This platform has a cost and sale price that will make it more than competitive in the highest volume, lowest price markets in the world, thus making a $100, feature rich, entry level smart phone a reality. This "cheap, but good" strategy fulfills the requirements of the above second example of doing the difficult things first.
Tackling the billion people population centers like China and India first is smart for Intel, because anything after that will seem easy. Keep in mind that China Mobile has more cell phone subscribers than the U.S. has people.
Rumors are that we will see the Intel high end smart phone solution next month at the World Mobility Conference.
It is becoming clear that Intel now has the leading solutions for their legacy businesses of PC, severs, and super computers, and competitive to leading solutions for the emerging markets of tablets and both low and high end smart phones. Is there any other company that has this kind of leverage on the worldwide computing market?
Those three underutilized Intel fabs will soon be producing low price, but high margin, entry level smart phone chips.
Let's not forget the automotive infotainment business that Intel is targeting. And what is this Intel Television thing all about? Apparently the intent is to trump Apple (AAPL) TV. I don't have a clue what they have in mind here, but the recurring thought I have is: Today Intel derives about $25 per year from the PC-using public ($100 CPU chip in a product that renews on a four year cycle.) Whatever this TV thing is would probably produce $25 worth of cost free service revenue to Intel each month on a similar sized using population. That's 12 times the current CPU revenue.
Since the blow-out numbers of 2009 and 2010, Intel has been assembling the assets to enable another sprint to a size that could be in the range of 2X the revenue of today.
Since that near-term growth has a very high probability of happening, I am using calls and LEAPs to play Intel. If you, as I am, are convinced of a doubling or tripling in the Intel stock price, it would be foolish to not use every bit of leverage the market allows. I'm buying INTC Jan 2014 30 calls for a dime to $.15.