Analysts revisit Nvidia stock price targets after surprise demand forecast

Nvidia CEO Jensen Huang's GTC address included a shocking outlook for AI computing demand.

Mar 19, 2025 - 12:38

Analysts revisit Nvidia stock price targets after surprise demand forecast

Nvidia shares bumped higher in early Wednesday trading, following the main event of the tech giant's developers' conference yesterday, as analysts dug into the keynote address from CEO Jensen Huang and his predictions for the AI investment landscape over the next few year.

Nvidia's (NVDA) investors appeared unmoved by Huang's presentation at the GTC event in San Jose yesterday, as well as the thirty-seven company press releases that followed it, even as the leading voice for AI advancements unveiled a new set of Blackwell processors and outlined a product development roadmap that underscores its unquestioned dominance in the market for the world's hottest technology.

Huang also insisted that the world "got it wrong" in their original assessment of DeepSeek's emergence earlier this year, arguing that the China-based startups AI success demonstrates the need for more, not less, computing power over the coming years.

"How do you find an algorithm whereby the more resource you provide, whatever the resource, the smarter the AI becomes," he asked. "The amount of computation we need at this point as a result of agentic AI as a result of reasoning is easily a hundred times more than we thought we needed this time last year."

Nvidia CEO Jensen Huang said the need for computing power is now "easily a hundred times more than we thought we needed this time last year."
Bloomberg/Getty Images

That need, Huang said, will likely require around $1 trillion of data center investments by 2028, a staggering tally that should underscore both demand for its new and existing line of AI processors as well as its ability to command higher prices and firmer profit margins.

Nvidia line-up extends to 2028

To meet that demand, Nvidia unveiled its new line of Blackwell 'ultra' processors, which are expected to ship later this year, with more than 1.5 times the computing power of the recently-released Blackwell.

Following on from that, Huang said, will be the late 2026 release of the next-generation Rubin line, which will come within a so-called server rack of 114 processors, double that of Blackwell, but with 3.3 times the power and capability.

A further iteration, Nvidia Feynman (named after the American theoretical physicist Richard Feynman) will arrive in 2028, Huang said.

"It is still Nvidia's game to lose - and they don’t appear to be losing," said Bernstein analyst Stacy Rasgon, who reiterated his $185 Nvidia price target following last night's GTC keynote.

"Nothing (in the speech) hugely surprised, given all the pre-event speculation, but we still thought it sounded good," Rasgon said. "The roadmap looks really solid, and their capability gap vs. competitors across their entire massive stack continues to widen."

Bank of America Vivek Arya, who held his $200 price target and 'buy' rating in place, also said the speech, as well as briefing from finance chief Colette Kress, "demonstrates Nvidia is continuing to deepen its competitive moat in a $1 trillion-plus infrastructure/services total addressable market."

"Overall, Nvidia continues to dominate the AI value chain with its full-stack turnkey (hardware, software, systems, services, developers) model," Arya said.

Computing power needs drive demand

Cantor Fitzgerald analyst C.J. Muse, who carries a $200 price target and 'overweight' rating on Nvidia stock, said the key element of Huang's address was the focus on computing needs and the company's ability to meet it with new products and applications.

"At the crux of everything here is Nvidia’s work in combining its best-in-class hardware with software initiatives, such as the newly announced Dynamo software stack that acts as the operating system for AI factories, and when combined with Blackwell offers ~40x inference improvement vs. the prior-generation Hopper," he said.

"Not to mention the roadmap the company laid out with the next 3 generations of product offerings, which bring along meaningful performance advantages," he added.

Wedbush analyst Dan Ives, meanwhile, focused on the impact Nvidia's advances will have across the broader tech sector, estimating that for every $1 spent on a Nvidia chip, 'there is a $8 to $10 multiplier across the tech ecosystem", including 'hyperscalers, software, data center buildouts, cyber security, and energy demand."

"In essence, Nvidia's chips remain the new oil or gold in this world for the tech ecosystem as there is only one chip in the world fueling this AI foundation - and its Nvidia," he added.

Financial follow-up

Nvidia plans to follow-up on Huang's address with a financial Q&A session later this morning at 11:30 am eastern time.

Last month, Nvidia topped Wall Street forecasts with fourth quarter revenues of $39.3 billion and operating income of around $25.52 billion.

Revenues from its key data center segment, which houses the chips and processors that power AI systems for clients such as Meta Platforms (META) , Microsoft (MSFT) , Amazon (AMZN) and Google parent Alphabet (GOOGL) , nearly doubled from last year to $94.5 billion.

More Nvidia:

Blackwell revenues, meanwhile, came in at $11 billion, a much stronger-than-expected total for the new line of processors that have been beset by reports of underwhelming performance and supply chain constraints.

Nvidia also said sees current-quarter revenues rising another 10% sequentially, to around $43 billion, amid what the company described as "rapidly rising demand for AI reasoning models and agents."

Nvidia shares were last marked 1.3% higher in premarket trading to indicate an opening bell price of $116.92 each.