A recent Seeking Alpha article looked at the scenario of Intel (INTC) buying Nvidia (NVDA). While it could provide some benefits to Intel, there is an even greater upside should Microsoft (MSFT) decide that it will completely mirror Apple's (AAPL) vertically integrated business model combined with its direct customer-selling channel. The PC market is shrinking while mobile continues to grow, and Microsoft has a rubicon to cross that is greater than selling tablets. Apple's introduction of its 64 bit A7 processor that is expected to soon show up in its A5 iPads is another indicator that it is coming after Microsoft's core business, which is corporate. In the short term, I expect it to counter with Intel-based product, however, it will not be sufficient in the long term as Intel is continuing to stick with its 60%+ gross margin model. If Microsoft buys Nvidia, it can eliminate a huge cost disadvantage across not just mobiles, but eventually PCs. The same argument, by the way, could also be made for Google (GOOG) buying Nvidia, because the Motorola acquisition did not go far enough.
Throughout the 30-year Wintel Empire, there have been many times when both Intel and Microsoft have tried to break the margins of the other by trying to create an alternative market segment with a new partner. In the early 1990s, Microsoft created Win NT with the purpose of supporting other server processor chips like DEC Alpha. In the late 1990s, Intel invested in Linux for not just the server market but also for desktops. The former worked, though the latter failed until Android appeared years later.
Going into the battle for mobile smartphones and tablets, the two have created a dual strategy of supporting each other and supporting the lower cost alternative. For Microsoft, it was to create Windows RT for the broader ARM mobile processor vendors. Meanwhile, Intel created an internal software team whose purpose was to tune Android for performance and low power on its Atom chips. Neither company has made dramatic improvements in the marketplace. In the case of Microsoft, customers balked at paying significant money for the Win RT operating system. And so the market split -- into cheap ARM based tablets running Android and expensive Apple iPads. Microsoft's entry with its own tablets has not gained the traction it expected, and they were considered expensive relative to Android tablets. In the corporate space, though, there is an overriding need to produce a hardware solution that is most likely capable of delivering the high margin Microsoft Office suite.
In an article yesterday in Venture Beat, Hewlett-Packard (HPQ) CEO Meg Whitman brought out into the open for the first time that HP views Microsoft and Intel as competitors. For many in the industry who have watched the market over the years, this was obvious. The margins of PC OEMs have been shrinking, however, they still exist and when added to the retail margin, they lose their competitiveness vis-à-vis Apple.
Apple's advantage in the mobile space is that its profit margin stack is considerably less than the combination of Wintel, the PC OEM and the channel. By building its own processors and O/S, Apple has only the foundry margin to pay and the rest is highly subjective as to what the market will bare. It is important to note that in the PC market, the elasticity is driven by a wide range of Intel processors selling at different price points as the rest of the hardware has been commoditized.
Microsoft has slowly been tracking Apple's moves. First it was the creation of stores and then next it was the selling of its own tablets. Following that, it was the acquisition of Nokia's (NOK) smartphone division. For Microsoft to be truly competitive, it has to make one more significant move and that is to build its own processor and then tune its operating system. Nvidia is the current ARM-based silicon provider in Microsoft's Surface tablets. The company is roughly $4B in revenue and sports an $8.8B market cap, but more importantly, it has the silicon and system expertise to broadly allow Microsoft to create lower cost mobiles and PCs. Nvidia's expertise could also be applied to its Xbox product line.
The compute market today is fracturing into two vectors with the high performance end becoming more and more data center centric and according to Intel, growing at 15% a year, while the client end, which is offloading more and more to the cloud, is commoditizing rapidly as it becomes more mobile. Intel averages a $100 CPU in the client space, although for many of the corporate ultrabooks, Intel charges over $200 for its ultra low power Haswell processor that weighs in at a die size 180mm. The dilemma here is that the Haswell line of processors should be downsizing to make ultrabooks more attractive, but Intel won't do that for fear of slashing revenues and therefore, Microsoft needs a lower cost alternative to go forward across the board.
Reading the HP and Microsoft tea leaves a little more closely, one could argue that Meg Whitman can foresee a time HP exits the PC business and hand it off completely to Microsoft, or perhaps should we say, Microsoft forces HP's hand. When Microsoft's OEM base consolidates or exits due to low or negative margins, then Redmond becomes the distribution channel and then it only needs to build the O/S for itself. If that is the case, then it can minimize its R&D by building the O/S to support just one underlying processor and it will not be built by Intel.
Microsoft knows that Apple is in the process of converting its complete line of computers starting with the iPhone, then the iPad and eventually the Mac to its 64-bit A7 and future chips that cost closer to $30 rather than Intel's $200 processors. If Microsoft is in the business of delivering a full software platform, then it needs to line up its hardware economics to compete. The quickest way to get started would be for Microsoft to buy NVdiia and recognize that it will have to deal with the repercussions from its existing customers and from Intel.