Sometimes, a company faces a huge transition, and in that transition, it faces challenges that, potentially, can mean a great loss of margin, profits, or even economic viability.
The products Nvidia (NVDA) sells, mostly GPUs (graphical processing units) and increasingly, mobile SoCs (System-on-a-Chip, marrying CPU, GPU, memory controller and, eventually, communications functions), are undergoing such a transition.
In the PC market, namely desktops and portables, where Nvidia sells mostly discrete GPUs, the GPU functionality is being increasingly built into the CPUs by both Intel (INTC) and Advanced Micro Devices (AMD).
While in the beginning discrete GPU cards had a huge advantage, where integrated graphics were clearly insufficient for most modern games, such an advantage is slipping away with each new integrated graphics generation, which increasingly is at least "good enough," pushing out discrete graphics toward much smaller niche markets.
(Click chart to expand)
What this means is that except for high-end niche segments, the NVDA product is becoming obsolete in this market. Now, looking at the latest 10-Q, one might think I am exaggerating since it shows NVDA growing the GPU segment from $581.8 million in revenue for the quarter ending on October 30, 2010, to $644.8 million for the most recent quarter, while also posting a tripling in operating income to $146.8 million. But the problem is, these numbers include licensing revenue from Intel, as part of a $1.5 billion agreement. This revenue, amounting to $66 million on this quarter, mean that the overall GPU revenue was actually flat. Also since licenses are all profit, the increase in operating income was much smaller.
The mobile market
The mobile market, which is cannibalizing the PC market, is also seen as the source of Nvidia's chance to grow. And indeed, right now NVDA is showing a powerful growth spurt from entering this market with its Tegra line of SoCs.
So, if this market is Nvidia's chance to grow, what's the problem with it? The problem is that the mobile market is much more competitive than the PC market, where NVDA formerly faced mostly a single competitor (AMD / ATI), whereas in the mobile market it faces almost half a dozen - Samsung, Qualcomm (QCOM), Texas Instruments (TXN), ST-Ericsson, Apple (AAPL) and, as of 2012, Intel, which recently seems to have broken into the mobile market.
Also, not only are the competitors much more numerous, but Nvidia doesn't control the main building block of its own CPUs, with these being based on ARM Holdings (ARMH) intellectual property.
Finally, while in the PC market Nvidia was, at times, a clear leader in GPU technology, this is not necessarily true in the mobile market, where the technological leader is said to be the Imagination Technologies PowerVR line, which powers the Apple iPhone 4S among others.
What this all means is that even though the mobile market for powering tablets and smartphones can present a revenue opportunity for NVDA, it's less clear that it will present a huge margin opportunity, due to much lower control of intellectual property and much higher competition.
Although NVDA's earnings will be somewhat distorted by Intel's $1.5 billion settlement over five years, NVDA seems to face significant headwinds both in the PC market, where each generation of integrated graphics encroaches more and more into its turf, and in the need to grow into a mobile device market where competition is much fiercer and margins are quite probably going to be lower.
Nvidia is not incredibly cheap while trading at a TTM EV/EBITDA of 7.4, right now its estimates imply no forward earnings growth and are suffering downward revisions - always a bad sign. While these conclusions might not be enough to warrant a short position, they seem enough to avoid the shares.