Wall Street has a new national pastime: seeing how many innings the AI boom can have.
Cantor’s $300 price hike and Morgan Stanley’s trillion-dollar forecast tell the same story: Nvidia is still running up the score in AI’s early innings

Samuel Boivin/NurPhoto via Getty Images
Wall Street has a new national pastime: seeing how many innings the AI boom can have.
Cantor Fitzgerald just raised its price target on Nvidia $NVDA to $300 (from $240), calling the company only “midgame” in the global race to build AI infrastructure, according to sites that saw the note. This is the most bullish call yet on a stock that’s already rewritten the record books — and a reminder that even trillion-dollar valuations haven’t satisfied investors’ appetite for GPUs.
The firm’s analysts reportedly expect Nvidia to post roughly $8 in earnings per share by 2026 and $11 by 2027, arguing that the chipmaker’s dominance in data centers, sovereign AI projects, and hyperscale partnerships has made its hardware less a product than a utility. The thinking is simple: Even if the hype fades, the infrastructure spending won’t.
According to Yahoo, analyst C.J. Muse, who holds an Overweight rating on the stock, said the company is still in the early stages of a multitrillion-dollar AI build-out that’s reshaping global infrastructure. He reportedly expects Nvidia to keep about 75% of the AI-accelerator market to itself, with a growth runway that stretches comfortably into the next decade.
Over at Morgan Stanley $MS, analysts reportedly came away from a meeting with Nvidia executives equally convinced that the runway is longer than it looks. In a client note viewed by others, the firm noted that most of Nvidia’s growth so far has come from cloud providers swapping CPUs for GPUs — a migration that’s still underway. The next wave, analysts said, will come from industries that haven’t yet gone through that shift: health care, legal services, industrial automation, and (eventually) robotics.
And that’s where the story gets bigger. Morgan Stanley reportedly estimates the total market for AI infrastructure could reach $3–5 trillion by 2030, as “new model categories” emerge to automate the physical world. That’s a bullish read not just on Nvidia, but on the idea that compute power itself is becoming the new capital stock.
Both firms reportedly waved off concerns about Nvidia’s vendor-financing push — the practice of helping partners pay for huge GPU orders — as a form of demand padding. Instead, they both essentially called it an accelerant: a way to speed up data-center capacity so new use cases can follow. In that line of thinking, the company’s tie-ups with CoreWeave and the U.K.’s sovereign investment fund are less about subsidies than about ensuring there’s always more compute to sell.
The chatter around Nvidia’s dominance can’t avoid AMD $AMD’s new deal with OpenAI. Morgan Stanley called Nvidia’s “full-stack” approach — its chips, software, and networking — the kind of ecosystem rivals can’t easily duplicate. Everyone else, they said, is still building around Nvidia, not against it.
Between Cantor’s record-high price target and Morgan Stanley’s insistence that the AI boom isn’t close to maturity, Nvidia has become Wall Street’s favorite Rorschach test.
So yes, Cantor is pushing a $300 target. Yes, Morgan Stanley thinks the AI boom barely shows up in the rearview mirror. These aren’t timid bets, but they speak to a shared confidence in a narrative that keeps expanding: Nvidia as the backbone of tomorrow, not just today. As long as that conviction holds, Nvidia might cross home plate before the rest of the market even picks up a bat. For a company perpetually stuck in the “early innings,” Nvidia seems awfully close to a title.
Join 500,000+ readers who start their day with Quartz.
By subscribing, you agree to our Terms of Service and Privacy Policy.