I, and apparently everyone else on Seeking Alpha, have an opinion on Intel (INTC) -- some favorable (me) and some not so favorable (the Grumps).
So, let's circle Intel from 30,000 feet and see what the business currently is and what it could be, based on fragments of information that we know and that is in the public domain.
I see six existing or potential major business operations for Intel:
Client (PCs that we humans use)
Servers (computers that computers use)
Foundry (making semiconductor devices for other companies)
Mobile (Tablets computers and smartphones, for the most part)
Security (It shelled out $7.8 billion for McAfee for SOME reason)
TV (Does anyone have a clue what this means?)
The Grumps like to hammer shareholders about the shrinkage and imminent demise of the traditional PC business. The reasons cited range from tablet cannibalization to smartphone cannibalization to even self-cannibalization in the form of Bay Trail eating into the iCore business.
I have a little different theory. I think that there is some level of self-inflicted Osborne Effect going on here. I believe, for reasons explained later in the article, that Intel feels an imperative to get to Haswell and future architectures and to 14nm Trigate and below.
Today the client business is the heart of Intel and will generate about two-thirds of the sales and, currently, one-third of the operating profit. In normal times, the client group performs more in line with the two-thirds representation in sales. Ramping new products always causes Intel margins to take a hit, and it seems like all it has done recently is ramp new products and processes.
Also the level of creativity on the part of PC OEMs compared to, say, Apple leaves something to be desired.
Well, computers need computers, and as you might suspect, a computer's computer is very powerful (and very expensive). The Data Center Group is 21% of the business and 36% of the operating profit.
Intel dominates the server business just as it dominates the client business. This server domination has not always been true. Intel worked very hard for many years on the server business, with the fruits of this effort only becoming visible in the 2009-2010 timeframe. Some of those little cuties have multi-thousand dollar price tags and 90% gross margins.
Intel is in the very early stages of establishing a foundry business that will be available to strategic, non-competitive companies. Some early examples are Cisco Systems (CSCO) and Altera (ALTR). There are some small sensor and start-up FPGA companies included in the foundry effort. Micro-Semi is also a foundry customer as well, for reasons that elude me. The conspiracy theorist in me thinks that Micro Semi is a conduit to get Intel processes into the hands of the military and other government "spook" programs without loading down Intel resources.
It is hard to imagine that, with the constraints involved, the foundry business ever exceeds $5 billion, ex Apple (AAPL).
Intel needs two blocks to play effectively in the mobile business. It needs a good application processor and a good communication solution. It has the former ready for tablet first (Bay Trail) and smartphones later (Merrifield). The communication block is a little more problematic. The newly released XMM7160 is OK, but not competitive with the best offered by Qualcomm. Intel will get both blocks right and be a major, if not dominant, player in the mobile business (I personally think it will be dominant in mobile as well).
For a minute I need to pass on my ideas about the size of the mobile market and specifically the smartphone market.
There are seven billion people on this rock. Two billion are too old or too young to know or care about smartphones. Another two billion are more concerned with what to eat today than a phone of any description. So, we have three billion people in developed and developing economies that might have use for a phone of some intelligence. Unlike the two-year renewal in the US, other parts of the world use a phone until it falls apart, and many of those are shared by whole families or villages.
Talk of high level smartphones in volumes of 1.5 billion per year is a tad high…. only my opinion, remember.
I think the Intel SERVED market (the market that could use what Intel has) for smartphones might be 750 million. No small number, but not 1.5 billion. If Intel were to win half of those units with an average price for the AP+Communications part at less than $15 (60+ gross margin on 14nm), the total business would be $5.6 billion. That is a big chunk of business, but is not impressive when compared to the $19 billion growth of 2010 and 2011 that had nothing (or everything) to do with Mobile considering that much of the growth came from the sever business, which "served" the mobile devices.
The conclusion one might come to here is that perhaps….just MAYBE, Intel does not see the smartphone business as such a critical business as the Intel critics (Grumps) might view it.
At the risk of incurring the wrath of the Grumps, I feel that the opportunity for Intel to grab most of the Apple "A" business is better than ever. The reason for that is that with the saving of Elpida, Samsung (SSNLF.PK) no longer has a stranglehold on Apple in the mobile DRAM area, and even the weak XMM7160 will make Qualcomm (QCOM) more cooperative in the LTE area. Apple is unique in that an integrated AP and LTE has no value for it since the AP is its own design and has no LTE on it. Just the Apple foundry win at about $30 for AP + LTE over 250 million units would produce $7.5 billion in revenue and make Intel a major mobile player overnight.
I have a difficult time believing that the start of the Intel Developers Conference and the Apple new product release are coincidentally on the same date, September 10. Could there be some joint something? It happened when the Mac went x86.
In 2010 Intel bought McAfee for more money than all of us could count in our lifetimes. Since that time McAfee has been selling its tired old anti-virus bundle as always and losing money for Intel. BTW, how can an established software company lose money?
The promise of Intel/McAfee has always been to link up the security expertise of McAfee with the silicon expertise of Intel in order to provide near perfect protection from those psychopaths who infect our computers and steal our data. I wrote about this a couple of weeks ago.
As I promised above, here is where, I think, the Intel imperative to get to Haswell and above, and 14nm and below comes from. Analysts have broiled Intel ever since the McAfee acquisition as an irresponsible wasting of shareholder money.
Intel, on the other hand, knew that it had to let Tick-Tock get it past Sandy Bridge and Ivy Bridge and into Haswell where the hooks necessary to fulfill the perfect security protection would be available.
Again we have to consider the market potential for perfect computer security protection. The LiveSafe security package from Intel/McAfee is subscription based for $79.99 per year. Intel, even in a disaster year, will ship 300 million CPU chips, each of which should be paired up with the McAfee package to be protected from the psychopaths among us.
Will Intel get 100% buy-in at 80 skins per year to protect 300 million computers? Not likely. I think an average security revenue, to Intel, of about $30 per year per chip is achievable. Since this software will only work optimally with Haswell and above computer chips, it will take 5 years, give or take, to turn over the whole PC fleet (approximately 1.5 billion worldwide). After that, Intel/McAfee simply updates the customers' subscriptions to the security-ware over the Internet and collects $79.99 from the 35% of computer owners who choose to do so. A 1.5 billion unit base of PCs times an average of $30 would produce a new annual revenue stream to Intel of $45 billion. Software business of this type generally carries a gross margin of 80% which would produce net margins of 30%+. That's about $2/share just in security.
The potential for the Intel security business is as big as all of Intel today and there is no reason that it will not happen.
Just maybe Intel feels more of a sense of urgency about security than it does about mobile and the Grumps just don't see it.
I don't have clue what Intel intends to bring to the consumer by way of TV programming, but lord knows that ANYTHING it does would have a high probability of being better than what we have today.
I can only guess that it will be silicon based as all its other products are. Perhaps Intel has perfected hardware data compression and de-compression to the point that not much bandwidth or memory is required to do streaming video.
Maybe it has come up with a way to control everything from one remote. That would be a novel idea. Then I could line up all my current "clickers" on a railroad track somewhere.
Intel sales and earnings could double based on a perfect execution of hardware/software security. It could add $5-6 billion in a well-executed mobile business, $5 billion in foundry, and another $7.5 billion in Apple business. TV - who knows? Further out, the Internet of things is another "killer app." Intel digital radio will help on that one.
Intel has growth paths ahead, some of which are not available to any other company today.