NVIDIA (NASDAQ:NVDA) followed a typically robust first-quarter earnings report with an impressive presentation to investors in which management pegged its total addressable market across all products at $70 billion, or about 10 times the graphics maker's $6.9 billion in revenue for fiscal 2017. The one-two combo of a great earnings report along with a rosy investor presentation has sent the stock up another 38% since the beginning of May.
Over the last several years, NVIDIA has shifted from simply selling chips to a value-added platform strategy where it spreads its expertise in graphics processing unit (GPU) technology to a variety of uses like deep learning and self-driving cars. Customers are finding NVIDIA's solutions for these markets valuable not only for the speed and power of the GPUs themselves but also for the range of software solutions NVIDIA provides to customize its GPUs for a given task. NVIDIA has basically turned the GPU into a software-based platform, which has led to an increase in gross margin from 51.4% in fiscal 2012 to 58.8% in fiscal 2017, fueling a 173% increase in earnings per share.
The company has a $30 billion addressable market opportunity in data centers coupled with an additional $8 billion in self-driving cars. NVIDIA not only has more room to grow revenue but can also expand margins further in the future as it continues to build on its lead in GPU technology with more investment and innovation.
NVIDIA got a big boost a year ago with the introduction of the DGX-1 supercomputer. This cemented NVIDIA's lead in the datacenter market, driving revenue growth of 145% to $830 million for the segment. It's on pace to have another outstanding year, given first-quarter growth of 186% year over year to $409 million.
The DGX-1 is a single aluminum box that packs the power of 250 servers in one unit and cost about 75% less than buying all of those servers individually. For data scientists, the DGX-1 sounds too good to be true, but it exemplifies the innovation and expertise NVIDIA has in the GPU space to bring a solution like this to market.
A server anchored by central processing units (CPUs) is no match for supercomputers like the DGX-1, which is powered by up to eight Tesla P100 GPUs. This is why companies are turning to NVIDIA.
GPUs have the speed and power to process mountains of data much faster than CPUs, and NVIDIA is the GPU pioneer. Its GPU technology is in several of the most well-known cloud services, including Amazon's Amazon Web Services, Alphabet's Google Cloud Platform, and Microsoft Azure.
NVIDIA's automotive segment revenue has grown nearly tenfold in less than five years to $487 million in fiscal 2017. While driverless vehicles that require some human assistance will begin rolling out this year, advanced driverless cars that require hardly any human assistance won't hit the road until 2020. Most of NVIDIA's $8 billion opportunity in self-driving cars is estimated to derive from the advanced segment of the market.
NVIDIA just recently announced a collaboration with Toyota Motor to launch an advanced AI vehicle by 2021, which along with several other major partnerships goes a long way toward solidifying NVIDIA's lead in the self-driving car market.
While PC gaming is not the hypergrowth opportunity of data centers and self-driving cars, it is still NVIDIA's bread-and-butter business and continues to be an attractive market to serve, given the recent growth of the segment.
The increase in production quality of games and the growth of e-sports naturally prods gamers to upgrade to high-end GPUs, as gaming enthusiasts want to play games with graphics settings maxed out for the best resolution and frame rate available. NVIDIA has historically delivered gamers the graphics power they want while positioning its GPU roster to allow for average selling prices to climb, which have increased 12% annually from fiscal 2013 through fiscal 2017. Higher average selling prices have been a contributor to NVIDIA's margin expansion lately, and with the ongoing upgrade cycle with respect to its high-end Pascal GPUs, margins still have room to move higher.