In recommending NVIDIA
In the short term, NVIDIA will remain a highly cyclical company roughly tied to the demand of the larger PC industry. That said, tailwinds from Microsoft's
NVIDIA might be an odd choice on the surface. While PC sales are rebounding steadily, the disproportionate growth of netbooks shows that consumers are increasingly willing to sacrifice power, instead seeing computers more as browsing tools. That's hardly the kind of shift that benefits NVIDIA's powerful stand-alone graphics cards.
However, as one hand taketh, an even more powerful hand giveth back. I'm recommending NVIDIA partially because it's an industry leader in a larger trend that could take years to develop -- but which promises to vastly reshape the computing landscape, and which has made so many Intel investors flush with riches.
Talkin' bout a revolution
NVIDIA's long been dependent on its graphics-card business to generate the lion's share of its profits. Consumers would install its cards to create glossy graphics for games on their PC's. If that sounds like a business with shaky long-term prospects, it is. However, the true power of graphics processors was obscured in their former "gaming-centric" incarnation.
In recent years, graphics cards have been shifting to more practical applications. While central processors remain ideally suited as the workhorse in basic computer processes, powerful common visual processes like rendering high-definition video, or editing graphics and photos, can benefit tremendously from the legion of processing cores on a graphics chip. Developers have taken note, and industry standards increasingly support computing languages that allow programs to access the graphics processor for better performance.
However, GPUs might be ready to step beyond performing workaday tasks, and into the realm of dedicated supercomputers. NVIDIA boats that its new Fermi architecture can deliver high-performance computing at "1/10th the cost and 1/20th the power of traditional CPU-only servers."
Fermi is the next evolution of a trend that's long been benefiting NVIDIA. Its Professional Solutions Business grew at a 22.6% compounded annual growth rate between January 2006 and January 2009. Customers found that high-end graphics cards could do much more than create stunning special effects for movies or games. They could also calculate complex equations for the financial industry, perform geological functions for the energy industry, and perform admirably in scientific research. It appeared that halcyon times had arrived for high-performance graphics computing.
Then the recession hit ... and IT budgets fell off a cliff, taking NVIDIA's Professional Solutions revenue with them.
Look out below!
Researcher Gartner estimated in July that IT spending would decrease at a 6% rate in 2009. That alone is a tough pill to swallow, but it's even worse for NVIDIA. Among IT budgets, computer hardware was hit the hardest, falling a steep 16.3% in 2009. While NVIDIA's high-performance computing solutions can save money and are increasingly used for varying purposes, buying into them is a dicey proposition for risk-averse IT and research departments with tight budgets.
Simply put, once IT spending comes back online, there will be an underappreciated pent-up demand for these neglected areas of spending. Look for NVIDIA's high-end cards to outperform expectations in coming years.
Other avenues for growth
It's also worth noting NVIDIA's attempts to branch into other areas that are complementary to its current expertise.
The company is forging ahead with multiple products that take advantage of the trend toward smaller mobile computing. Its Tegra chip, based on low-power Arm Holdings
Success isn't guaranteed for these products. Texas Instruments
NVIDIA's success isn't assured by any means. It faces major competition directly from Advanced Micro Devices
The efficiency and size benefits of integrating central and graphics processors remains to be seen. If it's a sizeable benefit, NVIDIA could find itself scrambling for a solution. Of the three main players in the graphics space, NVIDIA is the only company without a central processor component.
Then again, the production setbacks of Fusion, and Intel's difficulty developing its (now on hiatus) Larrabee chip demonstrate the difficulty in graphics-card development. NVIDIA may have history on its side when it scoffs at the efficiencies of combining central and graphics processors, but an investor would be wise to carefully monitor AMD and Intel's progress in this area.
In the end, I see NVIDIA as a sort of "call option" on the growth of graphics computing. The company has a sizeable lead in a market that has tremendous avenues for growth. If the benefits of GPU computing don't materialize to the extent expected, it's still a solid company with an underappreciated economic moat in the complex graphics space. In addition, its leadership in high-end graphics cards should continue to excel even if the consumer space becomes more challenging. The company wouldn't pay off with tremendous benefits, but it should continue to have a large market for its products.
However, if graphics computing does take off, the addressable market looks huge. Count NVIDIA out of your portfolio at your own risk.
Which is the best stock for 2010? See all 13 candidates here.