SOPA Images: Some of Nvidia's employees are believed to be billionaires
Nvidia will see a huge increase in revenue with its next-generation Blackwell GPUs, and by 2030, it will be worth $10 trillion.
That's according to Beth Kindig, a technology analyst at I/O Fund, who said Nvidia's business has huge defensive walls.
Kindig said Nvidia's next pillar of growth will be its CUDA software platform.
According to I/O Fund technology analyst Beth Kindig, Nvidia's stock price could rise 258% from current levels, giving the company a market capitalization of $10 trillion by 2030.
These gains come after Nvidia's shares have more than doubled so far this year and are set to more than triple in 2023 due to the rapidly expanding adoption of its H100 AI chips.
Nvidia is valued at $2.8 trillion, making it the world's third-largest company.
Kindig said Nvidia's next-generation Blackwell GPU chips, along with its CUDA software platform and expansion into the automotive market, will further drive significant growth for the chip maker.
Kindig predicts that the Blackwell GPUs will surpass their predecessor, the H100, and generate $200 billion in data center revenue by the end of Nvidia's fiscal 2026.
Blackwell will “power and enable trillion-plus large-scale language models, which is exactly what the big tech companies are aiming for. When you put these components together, you have a very large hardware and data center division. Next comes software. And then the third is automotive. There's a lot more to come. This is very early days for Nvidia,” Kindig told CNBC on Tuesday.
Kindig's long-term bullish stance on Nvidia stems from his belief that the company has an “impenetrable moat” around its GPU business.
Kindig said the total AI data center market is expected to grow to $400 billion by 2027 and $1 trillion by 2030, which he said will be captured primarily by Nvidia, not its biggest rivals AMD or Intel.
“NVIDIA will be a big part of that,” Kindig said, noting that this is primarily due to NVIDIA's software products being integrated into the company's hardware products.
“The CUDA software platform is a platform that developers learn on. Like with iOS, developers were developing apps for the iPhone, so people were locked into the iPhone. The same thing is happening with Nvidia, where AI engineers are learning the CUDA platform to program the GPU, so they're locked in. This combination is what I now call an impenetrable moat,” Kindig explained.
Finally, Kindig said the custom silicon AI chips that tech giants like Amazon and Alphabet are developing in-house will never compete directly with Nvidia.
The story continues
“They're not going to commercialize and sell chips like Nvidia is, so Nvidia has a way to go in that regard,” Kindig said.
Business Insider reported last week that Amazon has struggled over the past four years to develop an AI GPU chip that can compete with Nvidia.
Read the original article on Business Insider
