Have you ever stopped to think about which company quietly runs the modern world?

Most people would guess Apple. Some would say Tesla. But there’s another name that rarely shows up in daily conversation, even though it is now worth more than the GDP of many countries, including Russia and Australia.

That company is NVIDIA.

Its rise looks sudden. Almost unreal. A gaming chip company turning into a multi-trillion-dollar force at the center of artificial intelligence. But the story didn’t begin in a lab, a boardroom, or a venture capital office.

It began in a diner.

In 1993, three engineers, Jensen Huang, Chris Malachowsky, and Curtis Priem, met at a Denny’s in East San Jose. They didn’t have a polished plan. They didn’t have much money. What they had was time, coffee, and an idea that computers could do more than basic tasks.

Back then, personal computers were clunky. They didn’t even have speakers or microphones as standard. They were functional, not exciting. The founders saw one clear opportunity. Video games.

At the time, advanced 3D graphics were locked inside expensive workstations used by professionals. The trio wanted to bring that power into regular homes. Games were their “killer app”.

Jensen Huang wasn’t immediately sold. He had a stable job. A family. Responsibilities. But after enough conversations, he agreed to help figure out what kind of company this could become.

They named it NVIDIA after cycling through words that started with “N” and “V”, short for “Next Version”, before landing on the Latin word for envy.

They had $40,000 to their name.

When Jensen pitched the idea to investors, most weren’t impressed. One response stuck with him. “People don’t start companies to play games.” Even his mother suggested he just find a proper job.

Eventually, they got a meeting with Don Valentine of Sequoia Capital. Jensen has admitted he completely bombed the pitch. But Valentine invested anyway, partly based on a recommendation from Jensen’s former boss.

There was just one warning. “If you lose my money, I’ll kill you.”

No pressure.

NVIDIA almost didn’t survive long enough to matter.

Their first major product, the NV1 chip, failed badly. It used a graphics approach that Microsoft chose not to support. Overnight, NVIDIA was irrelevant. Staff were laid off. Morale collapsed. By 1997, the company had enough cash for just one more month of payroll.

That experience shaped an internal belief that still exists today. NVIDIA is always “30 days from going out of business”.

That fear pushed them to move fast. They built a new chip, the Riva 128. This time, it worked. It sold a million units in four months and pulled the company back from the edge.

Soon after, NVIDIA introduced something new. The GPU. The Graphics Processing Unit. With the GeForce 256, computers suddenly became much better at handling large amounts of data in parallel.

That change unlocked a major deal. Microsoft chose NVIDIA to power the original Xbox. The contract came with a $200 million advance. For a young company, it was massive.

For years after that, NVIDIA was seen as a gaming success story. Strong products. Loyal fans. Solid business.

But in 2006, NVIDIA made a move that didn’t look important at the time. They released a tool called CUDA. It allowed developers to use GPUs for tasks beyond graphics.

It didn’t explode overnight. In fact, it took nearly two decades to fully pay off.

Slowly, researchers at universities, including places like Stanford, realized that GPUs were perfect for deep learning. A form of AI that relies on repeating calculations at a massive scale.

Jensen Huang saw something others didn’t. He decided to push NVIDIA fully in this direction. The company invested heavily in AI research when it wasn’t obvious there would ever be a return.

It was a gamble.

When OpenAI was just getting started, Jensen personally delivered NVIDIA’s first and the world’s smallest AI supercomputer, the DGX, to them. 

And then the future happened. Chatbots appeared, image generators spread fast, and every company wanted an AI feature. All of it needed computing power at a scale most firms didn’t have. Data centers expanded. Orders surged.

NVIDIA was ready. Its chips fit the job. Its software was already in use.

Today, NVIDIA chips power systems used by OpenAI, Google, and Meta. Much of the AI boom rests on infrastructure NVIDIA built years earlier.

Without those chips, today’s AI tools would not exist in their current form.

But what are the repercussions of their meteoric rise?

As NVIDIA’s chips spread across data centers and research labs, new problems followed. Not the traditional ones, but the structural strain. AI-powered systems consumed a town’s worth of electricity in a day, and training a single large model required as much power as hundreds of homes use in a year.

A basic solution to this was to cool it with water, and many facilities rely on water-based cooling systems. This means millions of litres of water are being used every day, often in areas already facing water stress.

It shows how AI doesn’t always solve all our problems; it can bring some new ones.

Despite these rising problems, NVIDIA’s scale today is hard to ignore. Its valuation stands around $2.7 trillion. Nearly ten times Netflix. Roughly double Meta. It controls over 80 per cent of the GPU market. It employs fewer people than Apple or Intel but pays exceptionally well, with average salaries above $220,000.

If NVIDIA were a country, its economic output would rival or exceed nations like Spain, Turkey, or Saudi Arabia.

And it’s not slowing down.

NVIDIA is now backing humanoid robotics through companies like Figure AI, alongside investors such as Jeff Bezos. Jensen Huang has spoken about a future where $2 trillion worth of data centers are built just to support AI systems.

There is even talk of sentient AI. Machines that could rival or surpass human skills in certain areas within the next decade. Whether that future excites or unsettles you, NVIDIA is building the engine behind it.

Where This Leaves Us

It’s strange to think that a company founded by three engineers who didn’t know how to run a business is now one of the most powerful firms on the planet.

Jensen Huang still owns about 3.6 per cent of NVIDIA. That stake alone is worth roughly $64 billion today. Not bad for an idea sketched out in a Denny’s booth.

NVIDIA’s story isn’t just about success. It’s about patience. About betting early on technology that looked immature. About building tools quietly while others chased headlines.

As AI becomes more embedded in daily life, the companies shaping its foundation will matter as much as the products we see on screens.

Once, NVIDIA’s logo was background noise before a game started.

Now, it’s a signal of where technology is heading.