Artificial intelligence -- or at least the idea -- is nothing new. In the last few years, though, the decades-old phrase has gone from sci-fi fantasy (or sci-fi doomsday, depending on the writer) to an actual mainstream reality. And for good reason. AI is powering incredible operational changes, providing better insight on decision making, automating monotonous and soul-crushing tasks to free people's time for higher-level activity, and saving businesses money.

One of the pioneers of the movement is NVIDIA (NASDAQ:NVDA), and the company's third-quarter 2019 earnings call featured AI unlike any I can recall in recent memory. Though most investors think of NVIDIA as the graphics processor unit (GPU) designer for high-end video gaming (which it most certainly is), it's often forgotten that much of NVIDIA's current cutting-edge work is because of video games. That includes AI, which is quickly taking over as the leading driver of growth for the chip giant.

A robotic arm holding an NVIDIA chip.

Image source: NVIDIA.

What the numbers said

With the third quarter now in the books, NVIDIA is now nearly back to its cryptocurrency-fueled high-water mark from a year ago. Third-quarter revenue was down 5% year over year to $3.01 billion but surged 17% sequentially over the previous quarter. Adjusted earnings were likewise down 3% from the year prior but up 44% sequentially.

Of course, it was gaming -- which still makes up more than half of the revenue total -- that drove results. With the busy holiday shopping season here and nearly all video game developers having adopted ray-tracing technology, gaming laptops and PCs (as well as workstations for graphic artists and other creative professionals) are hot items. CEO Jensen Huang said he looks forward to helping the hundreds of millions of computer gamers around the globe upgrade their rigs in the years ahead.

Business Segment

Q3 FY 2020

Q2 FY 2020

Q3 FY 2019

QOQ Increase (Decrease)

YOY Increase (Decrease)

Gaming

$1.66 billion

$1.31 billion

$1.76 billion

27%

(6%)

Professional visualization

$324 million

$291 million

$305 million

11%

6%

Data center

$726 million

$655 million

$792 million

11%

(8%)

Auto

$162 million

$209 million

$172 million

(22%)

(6%)

OEM & other

$143 million

$111 million

$148 million

29%

(3%)

FY = fiscal year. QOQ = quarter over quarter. YOY = year over year. Data source: NVIDIA.

In other segment news, the promising auto business was weak during the quarter, but management said that half of that revenue is still legacy auto infotainment, which has been declining as the space has gotten crowded and commoditized. Crowded and commoditized is not NVIDIA's specialty. Autonomous-vehicle revenue continues to increase in the meantime, but this is a much longer game than what can be measured in just a couple quarters. Data centers also notched another year-over-year decline but grew 11% from the second quarter.

Why you should forget the numbers

With video games taking the spotlight as far as numbers go, what's all the hubbub about AI -- especially given that the data center and auto segments are responsible for most of the AI business? It all has to do with some recent developments driving future expectations, especially when it comes to conversational AI and the problems it helps solve (like call center and other customer service automation).

As Huang expounded on the third-quarter earnings call, Alphabet's Google has made some breakthroughs in natural-language understanding and real-time conversation. The AI models running the programming are incredibly complex -- Huang said they usually have 10 to 20 times, sometimes 100 times, more parameters than running an image and video. NVIDIA's GPUs are ideal at accelerating the compute time for artificial conversation in real time.

An NVIDIA EGX intelligent edge unit, a box with several GPUs and other chips inside.

An NVIDIA EGX edge computing unit. Image source: NVIDIA.

NVIDIA also made a slew of announcements in the last quarter. It launched its Intelligent Edge platform, which is already getting put to use in retail, manufacturing, and logistics by the likes of Walmart, Procter & Gamble, BMW, Samsung, and the U.S. Postal Service. It's also getting into mobile telecom, using its GPUs and AI to power new 5G networks.

Huang covered a lot of ground on the call, but it's a must read. NVIDIA's technology is helping fuel change across industries worth tens of trillions of dollars a year, so the expectation that AI will eventually itself become a multitrillion-dollar-a-year industry isn't so far-fetched. It's taking time for business results to materialize, but the revolution is happening, and NVIDIA's suite of GPUs and software solutions built on top of them is front and center. The continued revenue recovery in the third quarter was nice, but this is a stock worth keeping as a core portfolio holding for the long haul.