Nvidia has hit new highs on the stock market in recent weeks, driven by optimism about the future of artificial intelligence (AI).
The chipmaker, a leading maker of graphics processing units (GPUs) typically used in AI, closed above $2 trillion for the first time on Friday, making it the third most valuable company on Wall Street after Microsoft and Apple. It became.
Given the vast moat that Nvidia has secured through years of investment and the development of its own widely used software ecosystem, it's unlikely competitors will be able to cross the gap anytime soon, experts say. says.
“I think this is a game Nvidia should lose, and right now they don't look like they're going to lose,” Stacey Rasgon, senior analyst at Bernstein Research, told The Hill.
Nvidia has been developing GPUs for decades. Tianqi Chen, an assistant professor in Carnegie Mellon University's machine learning department, said the chips were primarily used in video games until a discovery a decade ago prompted the machine learning community to start using GPUs.
Jeffrey Hinton, a computer scientist known as one of the “godfathers” of AI, discovered that GPUs were more efficient at the large-scale computing required for machine learning, starting the “deep learning revolution.” According to Chen, this led to He said.
“The machine learning community has started to embrace GPU computing,” Chen says. “As of today, almost all AI models that run deep neural learning networks…many, even most, of them run on his GPU.”
Nvidia noticed this development and began creating a library for machine learning called CUDA within its software ecosystem, Rasgon said.
While Nvidia has focused on developing AI capabilities, its main competitor in the GPU market, Advanced Micro Devices (AMD), has struggled.
“It's only in the last few years that even AMD and other companies have been able to line up the resources to start investing in AI with data centers and GPUs,” Rasgon said. “But by that time he had led Nvidia by about 10 years.”
“A lot of it comes down to this. They recognized early on that this was going to be important. They put resources into product development, both hardware and software, towards this. They had competitors who were either not interested in this or incapable of doing it. And they never lost faith, ”he added.
Nvidia's significant lead in GPUs is strengthened by the presence of its own software ecosystem.
“Nvidia was an early pioneer in AI GPU hardware, but more importantly, we developed our own software platform, Cuda. These tools enable AI developers to build models on Nvidia. ,” Morningstar equity strategist Brian Colello said in a recent report.
“Not only does Nvidia lead in hardware, but it also benefits from high customer switching costs around Cuda, making it unlikely that another GPU vendor will emerge as the leader in AI training. thinking about.”
Rasgon pointed out that if developers try to switch to AMD or Intel parts, they will have to completely rewrite their code.
“It's a big undertaking,” he said. “And time is money, right? I mean, we want to get this product to market as quickly as possible. If we've been developing everything for the past 10 years with Nvidia parts, it's much more likely that we'll continue to use them.” It’s easy.”
NVIDIA has seen its GPU investments pay off over the past year, since the launch of OpenAI's popular ChatGPT tool sparked fierce competition among major technology companies to develop and release their own generative AI models. I'm starting to think I'm starting.
The chipmaker achieved a market value of $1 trillion for the first time in May 2023. NVIDIA has been on the rise in recent months, with the stock up 77% since the beginning of the year.
Nvidia initially crossed the $2 trillion mark late last month after reporting stronger-than-expected fourth-quarter results. Its market capitalization increased by $277 billion in one day, exceeding $2 trillion at one point, breaking Wall Street's record for the largest single-day gain.
Nvidia stock soared again last week, closing above $2 trillion for the first time after Dell reported better-than-expected fourth-quarter results. According to Reuters, Dell uses Nvidia GPUs in its servers.
“Accelerated computing and generative AI are reaching a tipping point. Demand is exploding across companies, industries, and nations around the world,” said Jensen, Founder and CEO of NVIDIA. Mr. Hwang said this in the company's latest financial report.
The company took a hit in China last year, with its data center revenue in the region declining “significantly” after the Biden administration imposed limits on advanced chip exports in late 2022.
Nvidia developed two new chips with reduced functionality to circumvent the restrictions, but the administration ultimately declined to do so, citing concerns that U.S. technology could be used to strengthen China's military. In October 2023, these chips were also cracked down.
The Santa Clara, Calif.-based company's dominance appears to be little shaken, at least for now.
Peer chipmakers like AMD and Intel are unlikely to move ahead of Nvidia, and while in-house options from companies like Google, Microsoft, Amazon, and Meta may be useful for certain purposes, they lack the flexibility offered by GPUs. Rasgon said it was likely.
Chen suggested that alternative chips could gain some market share if companies invest heavily in software components. But he also said he doesn't think Nvidia will lose its leadership position.
“Longer term, we expect big tech companies to strive to find second sources or in-house solutions to diversify away from Nvidia in the AI space, but perhaps these efforts will chip away at Nvidia’s AI advantage. But it will not replace it,” Morningstar's Collero added.
Copyright 2024 Nexstar Media Inc. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.