Any discussion about Intel Corporation (INTC) these days seems to begin and end with some form of "Intel missed (blew, failed to execute, ignored) the mobile business." There is a kernel of truth in that opinion if, and only if, your definition of "mobile business" is defined as smartphone and tablet SoC (System on a Chip) chips.
If you zoom back from that narrow view, it becomes clear that Intel Corporation has led the mobility trend beginning with the Centrino platform of nearly 10 years ago. Centrino brought standardization and reliability to the mobile PC experience.
During the past five years, Intel has been engaged in an effort to gain control of the data center server chip business. Some of the competition for the server business at that time was Oracle Corporation (ORCL) SPARC, IBM (IBM) Power, and Advanced Micro Devices (AMD) in x86.
The current status of the server market is: SPARC is being put out to pasture. The IBM Power chip is still being used, but is losing ground rapidly to x86. Even IBM offers x86 server products. IBM will continue with the Power architecture because of corporate ego and it needs to fill the fabs it has.
Apple's (AAPL) move from the Power PC chip to Intel x86 in 2006 was because of lack of roadmap improvements for the Power architecture. Advanced Micro Devices had some promise, but has since become a "dead man walking."
The Intel x86 architecture has clearly won the server segment, and within the x86, Intel has 95% market share and AMD has 5% market share and is declining. All other server architectures have about 10% market share that is rapidly moving to x86. This is a huge win for Intel. Being the dominant architecture to power the cloud is certainly "mobile" business. You just can't carry it in your pocket.
The high performance computing (super computer) market is an extension of the technology that has won the server business. Intel's Knights Corner chip has pretty much shut down all contenders in high performance computing. Intel needed to consolidate and control the server and HPC segments, since it is a contest that will be won once and never fought again, due to the R&D and capital requirements to be a player.
The hundreds of millions of PC CPU chips sold by Intel has provided the necessary scale to win a dominant position in the server and HPC segment. Servers and HPC are a $12-15 billion business that is essentially sole sourced to Intel today. The company simply had to bolt down these segments before moving on to smartphone and tablet SoCs.
For the tech-challenged pundits who think that RISC trumps CISC, I might point out that SPARC and Power are both RISC architectures that will fall victim to the CISC x86 architecture.
To summarize, Intel has a near monopoly on all computing CPUs from $30 entry level Atom chips through $4,000 server chips, and on to $10,000 Knights Corner chips.
What's left? The answer is the $10-$15 billion smartphone and tablet application processor market. Much of the Intel resources that have been leveraged to control everything above these "pocket mobile" devices is now available to target smartphone chips.
The Intel Medfield SoC has found some non-US smartphone designs. To get a little peek at what faces Nvidia Corporation (NVDA) and Qualcomm (QCOM), read this very slowly. The author has found a prototype of a Clovertrail powered tablet to benchmark against the Microsoft Surface and the ASUS VivoTab, both of which are powered by Nvidia Tegra 3 application processors. The Clovertrail tablet makes the other two look like toys.
The Clovertrail is basically a tweaked dual core Medfield. The Clovertrail is still made on the 32nm process. We can only speculate about performance of the out-of-order next generation Atom core built on 22nm and 14nm, that will ramp during 2013.
According to Intel, a 4G LTE function is coming by the end of 2012 and will ramp quickly in 2013. It is unclear whether it will be a stand-alone device or be embedded in a quad core Atom based SoC, or both. In any case, it will be a bad dream for Qualcomm.
There will be a "forehead slap" of global proportions when this Intel story comes into focus for analysts and institutional investors.
Better to buy Intel now with the knowledge that it is unavoidable that Intel dominates smartphone chips, just as it dominates all higher levels of computing devices.