The basic thesis of the Intel (INTC) bears is that Intel has failed to capture the increasingly important smartphone and tablet mobile markets and will die out with PCs and hard drive makers over time. The basic thesis of the Intel bulls is that its lead in process technology (used to actually make semiconductor chips) and overall IP will allow it to enter these mobile markets and eventually grab the bulk of the market share and profits. A secondary bull thesis is that the PC and server market is not as dead as the bears think.
My sense is that the bears are winning the argument so far, given the obvious contrast between Intel's cash position, earnings and dividend and its stock price result in a low trailing P/E ratio of 12. I don't see a lot of downside if the bears are right: it would take years, maybe over a decade, for the PC market to shrink to the point where Intel would be unable to generate earnings or pay a dividend. It is the upside potential, if the bulls are right, that I think is worth consideration.
Why would Intel be able to capture significant market share in mobile (tablet and smartphone, not netbook or notebook) in 2014? The thesis boils down to: process technology advantages will result in new power-efficient chips that take share from competition. The theory is that smaller transistors, in particular 3-D transistors, will triumph.
There is no arguing that in general terms Intel has a substantial lead in process technology. The problem with the argument is that Intel has always had a substantial lead in process technology. So they have had a lead in process technology over the past 10 years. It did not get them much smartphone/tablet business yet. We need to know why not, if we believe something has changed that will have a positive impact.
In process technology simply shrinking nanometers has not proven to be a universal solution. We know that when AMD (AMD) bought graphics-chip maker ATI it proved to be difficult to combine graphics technology and x86 CPU technology on the same die. Fully integrated smartphone chips have even greater integration issues. They must include analog circuits for wireless capabilities.
Chips for smartphones are much smaller than chips for PCs and servers. Being at 22 nm when AMD is at 28 nm is a big CPU advantage for Intel, but it is a much more minor advantage in smartphone chips. The reality is that Qualcomm (QCOM), TI (TXN), Samsung, Nvidia (NVDA) and even Marvel (MRVL) (which bought Intel's XScale ARM-architecture cellphone chip division in 2006) are well ahead of Intel in analog wireless technology and in smartphone application processor market share. The ARM architecture offers advantages, including software stacks, that x86 can't duplicate just by shrinking (just as x86 does things well that ARM can't do well by simply expanding). If it doesn't work well, making it smaller won't make it work better. In retrospect Intel would have been more competitive if it had not given up on XScale.
The bullish idea that Taiwan Semiconductor (TSM) and other foundries will have problems transitioning to 20 nm but Intel won't have problems transitioning to 14 nm is, at best, a guess. There are always problems going to newer process technologies. The only question is how long it takes scientists and engineers to overcome enough of the problems to get a commercial product yield.
Nevertheless, unlike smaller competitors, Intel is in a position to keep pushing into markets, if they choose. Their 22nm Haswell integrated graphics seem to be a significant leap forward (taking Intel's word for it), though still not up to the level of AMD and Nvidia, and not guaranteeing efficiency in the smartphone graphics arena. They can throw engineers at the wireless problem. They may be able to disrupt established relationships between ARM-based smartphone chip makers and the smartphone OEMs by their well known methods of throwing money at partners they want.
In summary, Intel is not out of the competition, but they have no clear path to victory and it will continue to be expensive to establish a presence in smartphones. I suspect they will do better in tablets, sooner, because they can always use a discrete wireless chip in their designs. Windows 8.1 tablets running on x86 processors might finally make a dent in ARM dominance of low-power devices.
Margins for these new products, I would hazard, won't be that great. The reason Intel is sitting on loads of cash is that in the x86 CPU space AMD has not been able to seriously compete in high-end PCs or servers. That leaves Intel able to set prices. They won't be able to set prices in the mobile space. Qualcomm, in particular, is well-funded if it gets to be a battle over market share.
I think the bearish idea that the PC era is over is wrong, too. We are seeing a shift, but by the time tablets can really replace a PC, the chips inside will have PC-like capabilities. They are as likely to be x86 chips as ARM-based chips. I think we are 4 years or so away from being able to plug a power-efficient ARM-based tablet into a large screen monitor and get good, PC-like results. By then both AMD and Intel chips will be power-efficient enough to serve in either non-mobile computers (PCs) or mobile (tablets).
ARM may have a place in the server space, but it has some of the same issues in that space that x86 has in the smartphone space. You can't just connect an ARM chip to a high-speed network and call it a server. It requires a software and hardware infrastructure that will require time to become competitive.
Think about your Linux desktop. Don't have a Linux desktop? Funny, some pundits predicted five years ago (or was it ten?) we would all have them by now. Linux is a fine operating system, and it has taken over much of the cloud server space, and gave birth to Android, but not every predicted trend comes true.
My baseline estimate for Intel in 2014 is: about the same as 2012 and 2013, with more upside potential than downside risk. Partly the sheer size of Intel means that even if they take a few points more market share from AMD and pick up a bit of tablet share, it will hardly move the needle. A broader global upturn in personal and mobile computers sales would provide a more significant upside.
All this makes Intel a good solid buy at this price (closing at $24.07 on May 22, 2013). The dividend of 3.7% should be attractive to conservative investors.
I am personally not willing to trade out one of (what I believe) to be one of the higher-alpha stocks in my portfolio for Intel at this point. I tend to buy stocks with smaller market capitalizations. I am willing to buy large-cap stocks where the market got the story wrong (I bought Seagate (STX) at $23.13, it closed at $42.17 today), but Intel's stock price is not that out of skew at this point.