I attended NVIDIA's (NVDA) shareholder meeting on June 19, 2008. NVIDIA has an Asian CEO, Jen-Hsun Huang, who comes across as knowledgable and enthusiastic. In San Jose, where so many Southeast Asians, Middle Easterners and Asians drive innovation, most companies still have no diversity in upper management ranks. While NVIDIA appears to be diverse, it's not helping the company's share price, which is down significantly this year.
NVIDIA designs graphics chips and cards. NVIDIA has partner fabs that actually make the chips which are then sent to partner factories to be mounted on video cards for PCs. NVIDIA graphics cards that bear NVIDIA as the manufacturer are referred to as reference cards.
NVDA's most touted product is its GeForce, a video graphics chip (NVDA likes to brag that the "graphics chip was given a soul" by its company). GeForce revenue increased 28%, and overall revenue increased 34%. Revenue numbers looked great, so it is quite possible the stock price is undervalued.
NVDA specializes in the high growth area of gaming, which requires high-end video graphics. It has even partnered with Hollywood movie companies to make films--the higher the graphics requirements, the more NVDA's products will be in demand.
As for competition, I asked my engineering friends why Intel wouldn't muscle in on NVDA's territory if NVDA is basically a fabless semiconductor company. Here is what one engineer said:
Intel did try to make a top-of-line video card at one time to compete with NVIDIA and ATI (AMD) but they failed. Modern day top-of-line video cards that NVIDIA and ATI (AMD) produce are very complex devices that take years of R&D. Video card drivers are notoriously complex pieces of software that also take years to fine-tune. Video card drivers support many if not all of the features of OpenGL and DirectX.
This anti-Intel piece brought some indignation from an Intel engineer, who referred everyone to "Larrabee." The Larrabee will not be available until Q1 2009, so until then, NVDA will be king of the graphics processing unit (GPU).
The highlight of the shareholder meeting was the "annual yearbook" video. It showed several fun pictures of the company and its worldwide employees, including a great Fortune cover with the CEO. It became apparent that NVDA's decision to open up its architecture through CUDA (a set of development tools) is something that should be analyzed in more depth. Like Linux, opening up a technology company's architecture allows hardcore geeks to improve your products for free, but this "free" work comes at a cost, namely less control over the product. Last time I checked, Microsoft's operating system is still king, despite the programming community's swearing that Linux's OS is superior. In any case, several images of "CUDA Everywhere" flashed on the screen.
The employee tidbits in the video yearbook were the most amusing--the video alternated between employees who had been there a few weeks to some who had been there for fifteen years.
After the video yearbook, which was impressively done, the presentation mentioned Quadro (Creativity), Tesla (Discover), and other NVDA products. The CEO stated his goal of making the 2-D into 3-D. Mr. Huang talked about four areas of innovation: photorealism; parallell processing; dimensionalization; and imaging and sensing, all tools to make the 2-D into beautiful 3-D. Mr. Huang compared his company with Nike several times, meaning that he wanted NVDA to be partners with the best video gaming companies and the best gamers, similar to Nike and Jordan, for example.
One slide showed the insides of a GeForce chip, more specifically the GTX 200 chip. It has 1.4 billion transistors, 933 GigaFLOPS of processing power, 240 processing core, and more than 20,000 threads. I have no idea what that means, but it sounds damn impressive.
Mr. Huang mentioned that the world's leading physics simulator used his products, and then took us into a John Madden NFL video game, where a player named "Gibbs" (Washington Redskins fans in the house?), showed how NVDA technology is used in more everyday applications.
Mr. Huang clearly wanted to show that NVDA was more than just a gaming company--he showed that his products help speed up iPod movie downloads (from conversion to finish) through the BadaBoom application, and talked about a new project, Folding@home, a Stanford-NVDA project. Other NVDA chip/card uses involved film, and we saw a clip of Curse of the Golden Flower, which apparently used the NVDA Tegra chip to create its visual effects.
There were only two questions. One person asked about the criteria NVDA uses to give back to the community. Mr. Huang said that the NVIDIA Foundation was independent and involved in diverse areas with no set criteria. NVDA has assisted battered women, schools, and other causes.
I asked the second question, about some specific items in the 10K, focusing on a section on competition from Intel, and who NVDA's major customers were. Mr. Huang talked about "having the courage to take the risk" of innovating into novel areas, and "imagining 'there' five years before going there." He said that NVDA's top customers were Dell and HP.
All in all, a very interesting presentation, but one that could have been done in less time. Some shareholders were getting antsy after more than an hour. The food, by the way, was good--I had some juice and a breakfast burrito.
With respect to the overal market, last week's stock markets swooned, and the Dow fell more than 300 points on one day. Is this the bottom? For long term investors, it may make some sense to wade back into the market and pick up shares of financially sound companies, such as Intel, GE, and perhaps even Pfizer. But short term investors willing to take on some risk may benefit also. I bought 1000 shares of GE on Friday and plan on selling it before Thursday. I predict GE will hit 27.50 at some point next week. Some brave souls willing to wait may want to consider NVDA as well.
Update on June 30, 2008: I picked up another 100 shares of PNK.
The information on this site is provided for discussion purposes only and does not constitute investing recommendations. Under no circumstances does this information represent a recommendation to buy or sell securities or make any kind of an investment. You are responsible for your own due diligence.