The dreaded cyclical downturn in the chip industry has arrived.Specifically, consumers are hitting the brakes on new PCs and smartphones, which has hit Nvidiaof (Nvidia 0.07%) Empires built on GPUs (graphics processing units) are especially hard. Last quarter, sales of video games and other graphics applications halved from a year earlier.
Enterprise computing, though, is another story entirely. Right now, sales of actual GPUs are still driving a lot of Nvidia’s growth, and a nascent software business is starting to emerge as well. Could 2023 be the year Nvidia makes the jump from semiconductors to software?
Will Nvidia’s software revenue take off in 2023?
To be clear, Nvidia does not separate its revenue from software services from sales of actual semiconductor hardware. However, we can assume that the majority of sales will still come from sales of GPUs and GPU-powered computing systems for use in data centers. After all, Nvidia has long bundled free software with chip purchases, making it easier for video game enthusiasts and developers to get their new GPUs running as quickly as possible.
However, Nvidia already makes more money from software sales than many investors realize. For example, many Nvidia chips are used for extremely complex computing tasks that require advanced software programs to function. Many companies that use Nvidia hardware aren’t software technologists themselves, so have already started using some of Nvidia’s software engineering. Still, some of that software access may be a one-time purchase rather than the recurring software-as-a-service (SaaS) solutions that investors prefer.
As proof, during the third-quarter fiscal 2023 earnings call, CFO Collette Kress explained that Nvidia will begin to unbundle its large “data center” division. Not coincidentally, “Data Center” is by far the company’s largest division. It posted sales of more than $3.8 billion in the quarter (more than half of its $5.8 billion in total revenue). Kress said:
As the number and size of public cloud computing and Internet services companies deploying NVIDIA AI grows, our traditional definition of hyperscale will need to be expanded to convey different end market use cases. We will adjust our data center customer reviews accordingly. Other verticals such as automotive and energy are also driving growth through key workloads related to autonomous driving, high-performance computing, simulation and analytics.
In other words, “data center” has become too general a term for Nvidia’s largest revenue segment. Included in the $3.8 billion quarterly slate are sales of actual GPUs to data center operators, GPU “rentals” through cloud services known as infrastructure-as-a-service, or IaaS, engineering services, and software for operations. GPU. Starting with its next earnings report, expect Nvidia to provide more details about what its customers are actually doing with its silicon designs.
Start small, but how small?
Sales for Nvidia’s data center division are a bit opaque right now, but what about the other three major end markets “gaming,” “professional visualization,” and “automotive”? There’s also no specific information on software sales here, but Nvidia may have made a lot of money from recurring software revenue.
Let’s start with Nvidia’s first subscription business, GeForce Now (think of it as Netflix (NASDAQ: NFLX), but for video games). Last summer, Kress said GeForce Now had more than 20 million registered users worldwide. How much revenue does this generate? Here’s some guesswork: GeForce Now has a free tier (I’m assuming half of the 20 million subscribers subscribe to this), $9.99 a month for “Priority” passes (assuming an average of 9 million subscribers, or $90 million in revenue) ), and a premium pass for $19.99 per month (1 million subscribers, or $20 million in revenue).
If that’s just a slightly more accurate assumption, that means Nvidia could be pulling in more than $300 million per quarter. In the third quarter of fiscal 2023, video game revenue totaled $1.57 billion, so video game streaming will likely remain a small — but certainly not insignificant — source of sales, supplementing video game GPU sales.
Something similar might happen with the “Professional Visualization” section, which caters to video game and film content creators, engineers, architects, and more. Revenue was just $200 million last quarter, a sharp drop due to lower PC sales. However, Nvidia’s Omniverse Cloud subscription (for building 3D virtual worlds, games and movies or simulating real-world locations) is included in this unit.
Omniverse is free for personal use, but the basic enterprise package starts at $9,000 per year. Hundreds of companies are now using Omniverse, so it doesn’t take many teams to subscribe to the service for it to become a big chunk of the “professional visualization” segment.
The “car” part is trickier. With sales of $251 million last quarter, we can assume that most of the revenue came from sales of Nvidia computing systems (Nvidia DRIVE Orin and Thor chips) used in modern vehicles. Nvidia, however, offers companies developing self-driving car technology a full suite of software, from the actual computing chips to the software used to speed up development. However, many of these sales can be attributed to the “Data Center” segment. Look for more details starting in the fourth quarter of fiscal 2023.
Regardless, Nvidia has a lot going for it when it comes to software, and 2023 could be a big year. But even now, the company seems to have been quietly ramping up recurring software sales to complement its leading GPU designs. As more and more companies adopt Nvidia chips when developing their own artificial intelligence systems and 3D applications, there is huge potential for Nvidia software to continue to grow. This is a truly unique company operating in the vast semiconductor industry.