About a year ago, NVIDIA’s stock was trading a little under $30 per share. Today, it trades for over $100 per share. That’s a 250% jump, compared to a 19% increase for the S&P 500.
What’s going on?
In short, NVIDIA is using its top notch graphics processors to do a whole lot more than graphics. The company is moving squarely into Artificial Intelligence (AI), via the cloud and cars. And if you’re an investor, you love it.
How does NVIDA organize its business?
Before we even get to the money part, let’s look at how NVIDIA describes its own business. For reporting purposes, the company divides its products into two categories:
- Graphical processing units (GPUs)
- Tegra processors
As we’ll see, the GPUs form the lion’s share of NVIDIA’s business. Unsurprisingly, GPUs figure prominently in gaming. But NVIDIA is pushing its GPUs into a range of different applications, which is why investors are so excited.
Tegra processors are mobile processors that NVIDIA features in two classes of products: DRIVE and SHIELD. DRIVE is a way for automobiles to crunch all kinds of data. SHIELD is a family of portable devices meant for mobile gaming.
NVIDIA uses these two categories (GPUs and Tegra) of products to attack five different markets:
- Professional Visualization
- OEM & IP
Gaming is clear.
For Professional Visualization, think of scientists, engineers, and product designers. People who need to see a visual representation of what they’re working on. We’re often talking about complex visual renderings built from tons of data. Something you’d want a fancy graphics-dedicated processor to handle.
With Datacenter, we’re talking about high-performance computing. This is where we have large-scale simulations, from the astrophysical to the quantum mechanical. (With my background in computational fluid dynamics, this application area definitely captures my attention. I spent a lot of time building and managing our computational clusters in grad school.)
Automotive is what you’d expect. Given the push toward self-driving cars, we need more processing capacity. We need cars to absorb more input data, and to make better “decisions” in response to that data. The processing load is enormous here.
Finally, OEM & IP isn’t really a market, per se. This bucket captures sales to companies that integrate NVIDIA’s products (OEM) or technology (IP) into their own offerings. These companies can target a range of markets. Given its limited visibility into the final applications, NVIDIA simply books these under the OEM & IP label.
How does NVIDIA make money?
With a company, when we talk about “making money”, we’re really interested in two elements: revenue and earnings. Revenue is the money the company brings in via sales. Earnings, or profit, is the money the company keeps after it pays its costs.
Let’s look at NVIDIA’s revenue
We can look at NVIDIA’s revenue in three ways:
- By business segment
- By market
- By geographic region
The next three plots show these views. We’ll look at revenue from the quarter that closed at the end of October 2016. We’ll compare against the revenue from the same quarter one year ago.
The plot below shows us revenue by business segment. We can immediately make two observations. First, the GPU business swamps the Tegra Processor and All Other businesses. In the most recent quarter, the GPU business comprised 85% of revenue. Second, we see that the GPU and Tegra Processor businesses both grew aggressively, year over year. The GPU business grew 53%. The Tegra Processor business grew 87%.
The next plot shows us revenue by market. Here we get a more detailed look at the business, as opposed to just GPUs versus Tegra Processors. NVIDIA is best known for gaming. We see the revenue stream reflects that reality, since gaming made up 62% of the total quarterly revenue.
This plot also gives us more perspective about NVIDIA’s growth. The fact that gaming, NVIDIA’s largest market, grew by 63% year over year is incredibly impressive. Look at Datacenter. That market grew nearly 200%. Automotive grew 61%. NVIDIA shows growth across several markets, which gets investors really excited.
The third plot breaks quarterly revenue down by geographic region. We can see that NVIDIA’s performance is dominated by Asia, where it generated 70% of its revenue. The United States, for the massive market that it is, only accounts for 14% of NVIDIA’s quarterly revenue (at least of the most recent quarter).
Like with the market picture, we see growth across the board, geographically. It’s one thing to see growth in your smallest segments, or across your smallest markets. It’s easier to grow something that starts small. But growing your largest markets or regions? That’s incredible. And NVIDIA grew every single geographic region it touched, year over year.
Let’s look at NVIDIA’s profitability
Having looked at NVIDIA’s revenue, let’s turn to profitability. Investors ultimately care about earnings. Revenue is nice, but if it doesn’t translate to profit, who cares?
Unfortunately, NVIDIA only reports operating income for its business segments (GPU vs. Tegra Processor vs. All Others). We can see the detail in the chart below. The GPU business showed an 85% increase in operating income, year over year. The Tegra Processor business pulled into the black, moving from $65 million in operating losses to $17 million in operating income. The All Other business remained flat (in the red, unfortunately).
Let’s refer back to the first plot in this post (which I’ve reproduced below), showing revenue by segment. It’s helpful to look at margins, which we find by dividing profit dollars by revenue. In this case, since we’re looking at operating income (in dollars), we’ll look at operating margin (in percentages).
The GPU business became more profitable, improving from a 33% operating margin last year to a 40% operating margin in the most recent quarter. Another way to look at this progress is via “incrementals”, or the ratio of incremental operating income dollars to incremental revenue dollars. In this case, the GPU business had year-over-year incrementals of 53%.
[The GPU business had $311 million in incremental operating income(= $678 million — $367 million), against $587 million in incremental revenue (= $1,697 million — $1,110 million), which gives us 53% incrementals ( = $311 million / $587 million).]
The Tegra Processor business flourished, moving from a -50% operating margin to +7%. It had 73% year-over-year incrementals, which is impressive.
Finally, the All Other business was flat, margin-wise. It had similar revenue and operating income both quarters.
What’s fueling all this growth?
It’s one thing to see growth in revenue and earnings figures. It’s another thing to understand what’s actually happening.
Why is NVIDIA growing so quickly? And why do investors believe this growth is sustainable (and thus justifies the higher stock price)?
In preparing this post, I went back and read the transcript for NVIDIA’s most recent earnings call. It’s a gold mine. NVIDIA’s CEO, Jen-Hsun Huang, does a masterful job describing his business.
If I could choose one line of his, to explain the current excitement over NVIDIA, it would be this one:
“…the mega point though is really the size of the industries we’re now able to engage. In no time in the history of our company have we ever been able to engage industries of this magnitude. And so that’s the exciting part I think in the final analysis.”
You can use buzz words like “artificial intelligence”, or “deep learning”, or “machine learning”, or “virtual reality”, whatever else you’d like. What really matters is (a) the size of the opportunity and (b) your ability to capture the opportunity. Investors believe these opportunities are becoming meaningful. And NVIDIA’s financial results indicate it can compete.
NVIDIA used to just make cutting edge graphics cards that helped you play fancy games. Now, it’s expanding into the cloud and the car. It’s trying to live at the epicenter of some of today’s most important tech trends.
It’s way beyond making just an expensive card you can slide into your PC. It’s making hardware, and software development tools, that multi-billion dollar companies can use to reshape their industries. It really is that big of a deal…if the company can execute, both technically and strategically.
That all sounds great, but what happened specifically over the past year? As far as I can tell from reading the most recent 10-Q report, two big developments carried the day:
- NVIDIA released graphics cards with a brand new, high-performance architecture (Pascal), that sold very, very well
- Demand in the Datacenter market boomed, as companies with heavy computational needs came to appreciate the value of NVIDIA’s GPUs
What’s cool is that these two developments point to NVIDIA’s past and future strengths.
In the past, NVIDIA was always strong with graphics cards. Management knows it’s an arm race out there. Gamers want the best specs possible. The more horse power they have, the better they can play the most demanding games. NVIDIA has consistently answered the call there.
In the future, NVIDIA will use its graphical processing expertise to overcome all kinds of data challenges. It’s not just about ramping up frame rates on graphics cards. As CEO Jen-Hsun Huang says, “Our GPU is not a specific function thing anymore. It’s a general purpose parallel processor.” And the Datacenter demand explosion indicates he might be right.
NVIDIA is a great example of the power of technology plus storytelling
NVIDIA has world class technology, no doubt. But great technology isn’t enough. NVIDIA needs to tell the right story. It needs to explain to investors how it’ll succeed in the future. It needs to explain to customers why its products are so valuable.
It’s not just investors and customers that need to hear a story, though. It’s developers, too. Developers are a big battle ground between Apple and Google. Both companies try to recruit developers to make must have apps for their mobile platforms. The better developers you pull in, the stronger draw you have for customers.
It’s not much different for NVIDIA. It’s one thing to have super powerful hardware. It’s another thing to give customers the confidence that they can take full advantage of that fancy hardware.
Here’s how Jen-Hsun Huan said it on the earnings call:
First of all, the reason why I’ve been on the road for almost two months solid is because the request and the demand if you will from developers all over the world for a better understanding of GPU computing and getting access to our platform and learning about all of the various applications that GPUs can now accelerate.
That’s critical. NVIDIA makes the hardware. NVIDIA supplies the development kit, so customers can write their own software. There’s no one size fits all solution here. NVIDIA needs the support of developers, so NVIDIA’s hardware can address all kinds of different challenges.
That’s an often overlooked element of storytelling. Having great technology is a nice start. But if no one knows about it, or no one knows how to use it, the technology is worth little. That’s probably the thing that impressed me most about Jen-Hsun Huang. He’s a great storyteller, and he knows how to speak to different communities.
Use NVIDIA as an example for your own career development
You have incredible technical skills. Make sure your storytelling skills don’t fall far behind.
Every single technical challenge you face has a broader business implication. Make those connections. That’s the magic of industry. You don’t work on technical problems just because they’re interesting. You work on them because solving those problems creates real value. Someone somewhere is willing to pay for the value you create. Have an idea what the right story is in that regard.
Take it from Jen-Hsun Huang. He clearly has enormous technical expertise. At the same time, he can zoom out without dumbing it down too much. He can still speak precisely, even when the audience doesn’t share his technical vocabulary. He can take on different perspectives, from those of research engineers, to software developers, to financial analysts. It’s an incredibly valuable skill to have. And it’s something you can learn, with just a little bit of focused effort, day after day.