The numbers were pretty good. Revenue was up and Tegra™ finally started to get traction, more than 3 times up but there are some red lights. First their revenues are down YoY. Second, their GPU business is down YoY and last, but not least, their professional business revenue is more or less flat for the last quarter.
The professional segment numbers are particularly worrying, because even in Nvidia’s worst results they generated a positive cash flow, and the GPU numbers do not help them.
Their professional business is heavily dependent on their GPU business. It is their GPU chips coupled with optimized drivers that transform 200 USD gamer cards into 2,000 USD professional workstation cards. While the bottom market of the GPU cards isn’t a blockbuster in terms of profit making, they help dilute the costs of chip development.
This bottom market is shrinking. While Sandy Bridge chips provide unbearable gaming performance, they are good enough to the greater majority of market and the bar will rise again when Ivy Bridge arrives next year.
AMD will be even more aggressive in their integrated GPU push. Llano will offer almost mainstream graphics performance with their IGP to the users, which means light gaming at marginal cost for the masses. Also, the ability of Llano processors to operate in Crossfire modes with cheap cards will be a plus to gamers, which will be a further hit to Nvidia’s market strategy.
A business that one year ago had almost 800 million USD now generates 637 million, and one has to wonder what will be the size of this business next year. If the business shrinks too much, they may have trouble funding research two or three generations ahead. A 200 million USD business such as Nvidia’s professional cards cannot sustain such high profit margins if the GPU business keeps declining like that.
On the good side, Nvidia is very bullish about their tegra chip. While Kal El *might* be a hit, we tend to be wary of Jen Hsung Huang declarations because of a history of misleading investors in public conferences. We agree that Android tablets will take some time to get traction and we agree that 30 months is a reasonable timeframe.
We will be observing, cautiously, developments in the mobile market. If we assume that Nvidia’s GPU market will be done and gone in the next 5, which is the timeframe of three GPU generations, this means that Nvidia’s consumer division must grow six fold just to make up for the revenue losses, and the professional market must embrace their mobile developments to substitute the GPU developments. The next quarters will be very interesting for Nvidia.S|A
Follow these categories: Microprocessors