NVIDIA (NASDAQ: NVDA) is turning out AI factories in relentless race to stay on top

NVIDIA (NASDAQ: NVDA) is already the third most valuable company in the world, but CEO Jensen Huang says it’s just starting to scratch the surface of the ongoing AI revolution that’s being largely powered by the chips it makes. With competition nipping at its heels, however, just how long the company’s stock can keep delivering is a trillion-dollar question.

Amid sky-high expectations, Huang has used several recent investor presentations to depict the company as much more than a chipmaker, emphasizing its role in developing software and data-center design. It’s a smart move as software companies tend to receive higher valuations than their hardware-focused counterparts, but it’s not one without risks. Increased complexity can create more ways to make a mistake, and it’s a rare company that can stay on top, doing everything, forever.

The once mighty BlackBerry stands as a cautionary warning, having gone from remaking mobile messaging and dominating corporate communications to obscurity in only a few short years. Faced with strategic challenges, the company was unable to strike the right balance between changing hardware and software preferences, allowing competitors like Google (NASDAQ: GOOGL), with its open-source Android OS, and Apple (NASDAQ: AAPL), with the iPhone, to quickly out-innovate it. It’s perhaps that latter company founded by Steve Jobs which serves as a counter-example of getting the mix right, and NVIDIA—with its CUDA software layer that developers use to program its GPUs—is trying to take that narrow path. The company is leading the pack in terms of hardware development, and its sticky software is keeping customers embedded in the ecosystem.

As Huang sees it, NVIDIA has already become a much broader AI factory, whose real product is new frontiers. He argues that computing is going through two platform transitions at the same time, both of which are fueling demand for the company’s products and services. The first is a general shift toward accelerated computing as Moore’s Law—a well-known principle that predicts the doubling of computing power every two years—possibly comes to an end. The second involves human engineered software transitioning to AI-developed software.

“AI is not about a chip,” Huang said last week at a Goldman Sachs conference. “AI is about an infrastructure. Today’s computing is not ‘build a chip and people come buy your chips, put it into a computer,’ that’s really kind of 1990s.” His discourse took a turn toward the meta as he described how NVIDIA engineers are using AI to write the code to make the AI. It’s a brave new world that is already starting to create itself. 

“What NVIDIA is really good at, as it turns out, is creating new markets,” Huang continued, detailing growing interest from robotics engineers to biomedical professionals using the technology to aid discovery of new drugs. “If you look at the whole IT industry, up until now, we’ve been making instruments and tools that people use. For the very first time, we’re going to create skills that augment people.”

It’s that exponential potential that has delivered returns of 158% to shareholders over the past year, but future results may only be as good as the company’s ability to keep making the right jumps before its challengers get there. Tech giants from Apple to Google to Microsoft (NASDAQ: MSFT), meanwhile, have all faced antitrust scrutiny over efforts to keep customers within their ecosystems, and there are early signs NVIDIA could get similar attention. While Huang made a solid case for the company’s continued leadership role in the relentless pursuit to the next plateau, he hinted that the biggest risks to the company might not be technological.

“Demand is so great that delivery of our components and our technology and our infrastructure and software is really emotional for people because it directly affects their revenues…It’s very emotional, it’s really tense,” Huang said. “It’s fun to be inventing the next computer era. It’s fun to see all these amazing applications being created. It’s incredible to see robots walking around. It’s amazing to see the AIs that we’re using to design the chips that will run our AIs. The part of it that is just really intense is just the world on our shoulders. And so less sleep is fine and three solid hours, that’s all we need.”

NVIDIA is clearly still on the vanguard with its next-generation Blackwell platform—described as a “step-function leap” over its current Hopper model—set to begin shipping later this year, but competitors like AMD (NASDAQ: AMD) are already making big moves, eager to challenge its dominance. Whether or not the company can keep up with its frenzied pace is anyone’s guess, but the real question might be just how long can Huang and his engineers go on getting by on three hours of sleep a night. The biggest challenge ahead might be profoundly human.