What passes for a computer is changing. It's changing toward tablets and smart phones, and one day not so far away even desktops and laptops/notebooks might start running iOS or Android, instead of the world evolving toward tablets and smart phones running Windows 8 and somehow needing Intel (INTC) processors.
It so happens that the desktops and portables running iOS and Android imply an incredible threat to Intel if they ever become reality. This is because these operating systems run mostly on ARM Holdings (ARMH)-based CPUs, and even if Intel was to successfully penetrate the mobile market, it might still make little difference because of the CPU pricing these architectures imply.
The pricing reason is simple. Whereas processors based on ARM IP (intellectual property) go for $5-$10-$15 (the whole chip), Intel is used to charging $40-$100-$300 for its wares. Indeed, estimates for the cost of the ARM IP only - what ARM holdings charges - run as low as $0.10-$0.30 per CPU, the rest being engineering, packaging and production on the side of whoever decides to use the IP. So, the world migrating toward the kind of market that exists today on the mobile CPU space could mean an incredible revenue and earnings downgrade for Intel.
One can speculate that Intel's chips are much more powerful than those based on ARM's technology. This is entirely true in what regards desktop and portable CPUs. But the problem is, ARM's CPUs are getting to be "good enough" for the majority of tasks, and it wouldn't be too far-fetched to think that Intel's wares could be relegated to a much smaller, niche, high-performance, market, much like what is slowly happening to discrete GPUs from nVidia (NVDA).
It should be said that the first attempt at such an evolution, the Google Chromebook, flopped mightily. This probably happened because of the lack of scale, consequent pricing that was comparable to full-fledged windows notebooks and diminished functionality. But this is probably not the last attempt made at bridging the gap between today's tablets and smart phones, and the PC market. And some future attempt might well succeed where Google failed, for instance the Asus Transformer Prime is already an example of a product bridging the gap and seemingly having some success. And even if these attempts mostly fail, there's usage substitution occurring even today. Many tasks that were performed in notebooks and laptops have now migrated to the tablet market, and that trend is only getting worse as the tablet market increases in size.
Also possible, along these lines, is the entry of ARM chips into the server market. Aside from the hardware, one of the largest costs with running a server farm is the electricity and cooling it demands. ARM chips, by being less power-hungry, could be reasonable choices in such an environment, even if they lose in absolute performance. Also, it's easy for these servers to run Linux, and much less critical that they run Windows. Both Hewlett Packard (HPQ) and Dell (DELL) are already able to ship these systems, though demand for them is still weak. This, however, might change during 2013 as ARM migrates to a 64-bit architecture with its ARMv8 design.
Several trends both ongoing (tablet substitution) and speculative (ARM entry into notebooks, laptops, desktops and servers) present a very real threat to Intel's market position. Even if these trends don't fully materialize, the emergence of $5-$10 CPUs using ARM IP doing the work of $50-$300 CPUs from Intel might at some point mean a considerable problem for Intel. It is likely that facing this possible development, Intel will continue trading at a discount to the market similar to the one it carries today (Intel trades at a forward 2012 P/E of just 11.4 times, and carries a juicy dividend yield of 3.1%).