trillion dollar company club.

It's made up of some of tech's biggest giants,

and for a moment, on May 30th, this company, Nvidia.

It's not a household name, but one day in late May,

Nvidia added nearly $184 billion in market value.

That's a single day sum worth more than other companies

in the semiconductor industry, and just a fraction

of its total value.

The company's record sales have been fueled

by an emerging technology.

- We're the world's engine for AI.

- [Narrator] Here's how an under-the-radar

graphics processor is cashing in on the AI craze

and what's next for its strategy.

- [Character] Excellent.

- [Narrator] When Nvidia was founded in 1993,

it established itself as a leader

in graphic processing units, or GPUs, for video games.

The company's powerful chips made images crisper

and less choppy.

- Nature of our technology

is to accelerate multimedia applications for the consumer.

- [Narrator] In the years since,

the company's chips have become an indispensable part

of many other products, from PCs to cars to robots.

Nvidia made a pivotal decision in 2006

when it started allowing software developers

to play with its GPUs.

- If you want to be a trillion dollar company,

you can't do just one thing, kind of like opening a door.

And once they opened that door, people started to tinker.

- [Narrator] One of those expansions

was cryptocurrency mining.

The company's GPUs could handle the large amount

of computational power the task required.

Business flourished.

Rising sales from the crypto rush

helped Nvidia pass rival Intel

in market value in 2020.

But sales fell as crypto crashed

and the gaming industry slumped.

Still, the chip maker was starting

to see another frontier boom: generative AI.

In May, sales of chips

for the data centers needed to power AI were up by 14%,

thanks in large part to the growing need

for Nvidia chips in AI calculations.

For Nvidia, that was a decade in the making.

The company's chips broke through in the AI community

in 2012 with a neural network called AlexNet.

- We saw early on about a decade or so ago

that this way of doing software could change everything.

- It was very futuristic.

It was very uncertain at that point.

Nobody knew how big AI was gonna be.

They made that big bet at a time when a lot

of other competitors were not making that kind of bet.

- We risked everything to pursue deep learning.

- [Narrator] Nvidia chips are suited for powering AI

because of their ability to do lots

of complex computations simultaneously.

Traditional processing units can't do that as quickly.

And other GPUs, like those from Intel and AMD,

haven't gained major market share in AI computing.

- If you think about how graphics work, your monitor,

your computer monitor has a lot of pixels,

thousands of pixels on it,

and your computer needs to calculate

at any given time what each

of those pixels need to look like.

It was discovered fairly quickly after Nvidia kind

of opened up the door for people to use its chips

for other things that they were actually really good

at the kind of mathematics

that underpins artificial intelligence.

- [Narrator] In late 2022,

AI exploded with the public release

of OpenAI's ChatGPT and Nvidia's stock skyrocketed.

While companies like Google, Microsoft,

and Amazon battle it out in the AI race,

Nvidia is raking in the cash.

That's because its chips power most of them.

- For Nvidia, it's a bonanza.

All of these folks who are competing to get a slice

of the generative AI pie need Nvidia.

- [Narrator] In 2018, the division that focuses

on AI computation made up less than half

of the company's sales.

By 2022, it was the majority.

- It went from Wall Street expectation of 7 billion

to Nvidia saying actually gonna make 11 billion.

And so that really opened people's eyes.

It was a real statement that, you know,

this generative AI craze that everybody's talking about

is more than just a hype cycle.

- [Narrator] But that success

is creating a different problem

for the company: supply.

- The demand is literally from every corner of the world.

- GPUs are at this point considerably harder

to get than drugs.

(Elon chuckling)

- [Narrator] The company's supply chain

hasn't been able to keep pace.

- The the chip manufacturing cycle is such

that you can't just immediately make more

of these things like today if you want them.

- The company says it has filled out some of its supply.

- [Colette] We are working on both supply today

for this quarter

but we have also procured a substantial amount

of supply for the second half.

- [Narrator] But the chip's global demand

has also caught the company

in the crosshairs of increasing tensions

between the US and China.

Last year, the company said $400 million

of its quarterly sales could be subject

to new US chip licensing requirements for China.

In response, Nvidia began offering an alternative chip

to customers in China.

Despite some roadblocks,

the chip industry has its eyes on Nvidia's success.

- They're still riding this boom.

It's no nowhere near done.

The real question is whether companies

can build huge, huge businesses that are based

on that artificial intelligence capability.

(pensive music)