SAN FRANCISCO — In high school, the only thing Pablo Nava Barrera knew about was Nvidia, which made the graphics card in his video game computer. But when he started looking for internships last fall, the second-year computer science student at San José State University discovered that the company’s chips were behind “generative” AI tools like ChatGPT and self-driving cars.
Now Nava, 19, not only has a summer internship at Nvidia, he’s also a shareholder. Last month, he used the money he had saved from gifts and part-time jobs to buy an undisclosed amount of Nvidia stock. He has already seen a gain of 15 percent.
Subscribe to The Post Most newsletter for the most important and interesting stories from The Washington Post.
“It’s impressive how they’ve gone from a company with the best graphics chips to an AI powerhouse,” says Nava, who plans to keep his investment “indefinitely.”
While the stock price has risen almost 300 percent in the past year, Nvidia fever has swept across US computer science and Silicon Valley. At college job fairs, hundreds of students line up to meet company recruiters. On TikTok and Reddit, people are debating whether it’s too late to invest. And along the highway between the valley and downtown San Francisco, billboards proclaim the availability of Nvidia’s flagship product — microprocessors known as GPUs — from startups that have bought the chips to meet overwhelming demand.
Nvidia’s chips are powering the AI models behind the artificial intelligence boom, including popular tools like ChatGPT. Big Tech companies, from Google to Meta to Amazon, need the chips to power their AI tools and big data centers. Demand is skyrocketing, pushing the price per chip to as much as $30,000 – if you can find one. The company is now worth more than $2 trillion – more than Google or Amazon – and is behind only Microsoft and Apple in the battle for the title of most valuable company in the world. This week, Nvidia’s stock price surge pushed the S&P 500 to a record close.
Top students who may have fought for a spot at Google or a flashy start-up in previous years are now applying to Nvidia. The company’s upcoming annual conference, which in recent years has been attended mainly by chip buyers from gadget manufacturers, has become a must-see event on the AI conference calendar. Google, Amazon and Facebook are buying more Nvidia chips for their own data centers to meet the massive – and growing – demands of AI processing.
Nvidia CEO Jensen Huang, who has the company’s logo tattooed on his left shoulder and wears a black leather jacket at public events, has become a god in tech and business circles, known simply as “Jensen” in bemused and envious tones named.
His trips to Taiwan, where Nvidia’s chips are manufactured and where he lived before emigrating to the United States with his parents as a child, are now media events. On a recent trip, locals recognized him in a tofu pudding shop at a night market and asked for selfies, prompting a TV news report.
The company’s meteoric rise is due to an “accident of history,” says Oren Etzioni, an AI professor at the University of Washington and founder of TrueMedia.org, a nonprofit AI deepfake detection organization. The chips Nvidia designed to run video game graphics turned out to be well-suited for performing the massive computations needed to train and run modern AI algorithms. Years before ChatGPT kicked off the AI boom in 2022, Nvidia began tweaking its chips and software to better serve AI researchers.
“It’s better to be lucky than smart. In this case, they were lucky, they were smart and they worked hard,” Etzioni said. “They’re riding an exponential wave, and that’s why you get to a $2 trillion company so quickly.”
A spokesperson for Nvidia declined to comment.
Nvidia was founded in 1993 in a Denny’s by Huang – who had been a chip designer at Advanced Micro Devices – former Sun Microsystems engineer Chris Malachowsky and former IBM engineer Curtis Priem. Video games were booming, and the three suspected that a new kind of chip, specialized for computer graphics, would be needed as games became more visually complex. The bet paid off and Nvidia steadily grew into one of the leading suppliers of GPUs, short for “graphics processing units.”
Around 2014, researchers working on AI realized that GPUs worked better for training AI than other chips. While the more common computer processing units, or CPUs, are optimized to perform a small number of very complicated tasks, GPUs are better at performing many simple calculations simultaneously. Because training AI requires making many connections between billions of different words or images, GPUs worked well. Many of the major AI breakthroughs that helped lay the foundation for the major language models at the heart of today’s AI revolution have come about running AI algorithms on Nvidia GPUs.
The company realized what was happening and doubled down on the AI industry. And as Big Tech executives began focusing their companies on AI, Nvidia’s revenues and stock price began to rise. By the time OpenAI made its ChatGPT chatbot public, kicking off the AI craze, Nvidia was already miles ahead.
“It’s a piece of technology that no other market has come close to in terms of functionality and reliability,” said Fred Havemeyer, an analyst at Macquarie Group.
Not only does Nvidia have the best chips, but the software it built to help program them is the most widely used among AI researchers and engineers. At this point, switching to another company’s chip would force them to learn a new software system.
In 2023, Nvidia sold about 78 percent of AI chips destined for data centers around the world, up from 63 percent in 2022, according to London research firm Omdia.
If AI were a gold rush, Nvidia would be the biggest pickaxe supplier, and experts say its dominance isn’t likely to wane anytime soon.
“The picture may look different in five years, but right now Nvidia is king,” Etzioni said.
– – –
Is it too late to come in?
When Nvidia published another record-breaking quarterly report last month, the excitement around the company was palpable on social media. In tweets and TikTok posts, former investors bitterly posted calculations of how much their shares would have been worth had they held on, while those who made money bragged about their gains and posted images from the movie “The Wolf of Wall Street.”
On Reddit threads about the stock, users debated whether it was too late to get in on the action.
“If you don’t see Nvidia’s opportunity right now, then you’re really lost,” one user wrote along with the facepalm emoji.
Kaitlin Mackie, director of brand design at digital healthcare company Welldoc, saw the impact AI image generators were having on her field and googled “How to invest in AI.” The internet led her to Nvidia and in November she bought 2.6 shares.
“I bought it because it was a company that drives the future, especially in my industry,” Mackie said in an interview.
On dates last year, men told her the stocks were overvalued, or wondered why she would bother owning so few shares in one company. She ignored them. It paid off when Nvidia reported earnings in late February and its holdings rose by $900. She used some of the money for lip filler and the rest for fundraising.
Her 3.96 shares of Nvidia are now valued at around $3,480, and she believes in the company.
“Even if it goes down again, I’m just going to hold on to it,” Mackie said.
Nvidia’s growth stands out among the general doom and gloom in Silicon Valley. While the rest of the industry has laid off tens of thousands of workers over the past two years, Nvidia’s workforce has grown 30 percent to 29,600, according to company documentation.
That growth has helped boost the economy in Santa Clara, a suburb sandwiched between San Jose and Cupertino, home of Apple. Some startups are moving their headquarters to the area to gain better access to Nvidia executives and sales representatives.
In February, a line of several hundred eager students gathered around a building on the Georgia Tech campus in Atlanta, all waiting to meet with visiting Nvidia engineers about internships and jobs at the company, according to Denitsa Dimitrova, an electrical engineering student who attended used to be.
Dimitrova, a senior, had never heard of the company before fall 2022, when a friend interned there.
“I didn’t think I would ever want to work there. Now they are a big name,” Dimitrova said. “Everyone is interested in being at the top of the top, and Nvidia has joined those ranks.”
On Handshake, a recruiting platform that connects students and employers, Nvidia internships received seven times as many applications in January as a year ago.
Silicon Valley companies have traditionally used the promise of stock growth as a key way to win over top prospects from colleges across the country. Over the past decade, thousands of young people have made their fortunes working for Big Tech companies and high-flying start-ups like Airbnb, Uber and Spotify.
But wave after wave of layoffs has damaged morale among Silicon Valley workers and shattered the perception that Big Tech provides job security. AI is the bright spot, with investors pouring huge sums of money into the space and AI researchers and engineers winning seven-figure compensation packages.
Even without the potential to make money from stock grants, Nvidia’s salaries are already as high or higher than those at Google, Facebook and Apple, the traditional top employers of the past decade, according to data from Glassdoor, which surveys tech workers.
Shiva Kumar Appam will graduate from San José State University in May with a master’s degree in electrical engineering. He applies to a few places, but Nvidia is one of his top choices.
“I would be on top of the world,” Appam, 25, said of the possibility of getting a job there. “It’s a dream company.”
– – –
Graphic:
https://washingtonpost.com/documents/26f09c00-e3c8-40fb-b729-062cdb7b117d.pdf
related content
Race is an ever-present source of tension in the Trump Georgia case
The Taliban once destroyed televisions. Now it encourages YouTubers to promote its image.
This agency is tasked with keeping AI safe. His offices are crumbling.