After a stellar run from the beginning of the year to about late April, semiconductor stocks have given up a good portion of their gains during May. Intel (INTC), in particular, was interesting because it was finally setting new multi-year highs after trading in a fairly narrow range.
Analysts upped their price targets to the low to mid $30s and all of a sudden people began to view Intel as a serious competitor to companies that dominate the ARM (ARMH) ecosystem such as Qualcomm (QCOM) and NVIDIA (NVDA). So, with the recent tech sector pullback, is now the time to go in and snag Intel shares at a discount? I believe the answer is a resounding yes!
First and foremost, Intel has an excellent dividend track record. It recently increased its dividend by 7 percent to 22.5¢ per share. As of the last close, this represents a 3.27% yield - which is respectable, especially since this exceeds the current yield of the 30-year U.S. treasury bond of 2.86%. For a company that is also the largest computer chip maker in the world with numerous growth opportunities, I'd say this is is a solid deal.
Now, let's talk about those growth opportunities. Intel's currently dominant in the PC client space, which consists of laptop and desktop PC, with over 80% of the market share there. However, Dell's (DELL) latest earnings call cast doubt on the growth of the PC space, citing a 12% drop in desktop PC revenue and a 13% drop in laptop revenue. So while it can be debated that this may be a Dell specific issue or not, it's clear that the PC segment is not the huge growth driver that it once was. Where's the growth going to come from, then? Mobile and the server/datacenter segment.
The mobile segment consisting of smartphones and tablets is always the headliner these days and it's an area that Intel's been late to address. It's why ARM commands a trailing P/E of 52, Qualcomm has a respectable trailing P/E of 20, and Intel is sitting in the corner an extremely conservative 10.87. Now, until Intel gave proof that it could design a viable smartphone SoC (system on chip), I could at least understand why Intel was trading at just shy of 11 times earnings - it was priced for stagnation. But earlier this year, Intel finally did it.
It designed an x86 based smartphone SoC called "Medfield" and a complete reference platform (that is, a phone design that other phone vendors could replicate and deploy if they so chose) that actually worked. Finally, an Intel-based phone was running the popular Google (GOOG) Android operating system.
In fact, shortly thereafter, an actual product, the Lava Xolo X900, based on Medfield and the Intel reference platform began shipping in India. The performance and battery life were extremely respectable, with the single Atom core in the Medfield SoC able to match or exceed the performance of dual core ARM Cortex A9 solutions almost across the board while keeping power consumption in check.
However, the quad core Cortex A9 implementations from Samsung and NVIDIA took the lead in applications that could use all four cores, and the Qualcomm Snapdragon S4, which features a custom-designed CPU core that runs the ARM instruction set, managed to edge out Medfield on a per-core basis while featuring two cores.
I'm not saying that Intel has this in the bag and that it'll be an easy march to capturing significant market share in the mobile sector. Medfield is already outperformed by Qualcomm's latest Snapdragon, and I suspect that most phones that run on ARM Cortex-A15 class chips will outperform the current Medfield offering.
But if Intel could take its 4-year-old Atom core (yes, the Atom core in Medfield is the same core, but die-shrunken and slightly tweaked for phones) and turn it into a competitive solution for today's smartphones, I am extremely excited to see Intel's new, redesigned "Silvermont" CPU and its successors (that will now come every two years instead of every five, thanks to its "tick-tock" cadence). With competitive mobile solutions and the ability to manufacture them in house, Intel will be able to significantly grow revenues, while keeping gross margins nice and high.
The next area of growth for Intel is in the server/datacenter. The path here is much more straightforward. Intel is the dominant force here and faces very little competition, especially as AMD's "Interlagos" offerings prove to be inferior to Intel's Xeon line in all metrics except, perhaps, cost and some really out there edge cases where AMD's brute number of weaker integer cores comes in handy. It leverages the CPU cores developed for the laptop and desktop segments, builds a server-grade system architecture around the cores, spends extra time validating them, and then deploy thems at a much more relaxed clip.
The growth here will come due to the ever expanding nature of the cloud. Phones and tablets can't store a whole huge amount of data, nor do they have the processing power to carry out extremely computationally heavy tasks. So this work is offloaded to servers, the vast majority of which utilize Intel Xeon processors (other players here include Oracle's (ORCL) SPARC processors and IBM's (IBM) Power processors). Server chips command extremely high margin compared to their client counterparts. Assuming the margins stay about the same, lack of competitive pressure on Intel in this segment should lead to material top and bottom line growth.
A caveat, however, to Intel's dominance in the high performance server CPU segment is the appearance of "micro-servers". The idea here is that a lot of tasks that require servers (such as hosting a website) don't actually do a whole lot of performance intensive computation and just need a high number of a fairly weak CPU cores and a lot of memory and storage space.
AMD, ARM, and Intel seem convinced that this is a nontrivial market, as evidenced by AMD's acquisition of SeaMicro, ARM's attempt to make its latest Cortex A-15 capable of addressing more than 4GB of memory and running at higher clock speeds, and Intel's release of Atom chips for just such uses.
It's not clear how much cannibalization the higher-end Xeon lineup will see due to the increase in micro-servers and it's not clear how much of a hit, if any, Intel will see on a gross margin front if microservers do in fact lead to such cannibalization. Intel does acknowledge the market and is releasing products to service it, but it pays to keep an eye on revenues and margins in this segment of Intel's business going forward.
To sum it all up, Intel's got a lot of interesting growth opportunities in the smartphone/tablet space going forward that make its current valuation look extremely cheap. Coupled with an excellent dividend yield sustained by its lucrative and steadily growing core PC chip business, I think at these levels Intel represents excellent value.
Disclosure: I have no positions in any stocks mentioned, but may initiate a long position in INTC over the next 72 hours.
