Nvidia shares surge 27% on historic $11B AI revenue forecast

If anyone was still debating whether the generative AI boom was just another tech bubble, Nvidia just put that argument to rest in historic fashion.

Following a blowout earnings call, Nvidia shares skyrocketed 27% on Thursday. If the gains hold, the chip giant will add more than $200 billion to its market capitalization in a single trading session, eclipsing Apple’s previous record of a $191 billion single-day gain set back in November 2022.

The company’s ticker quickly became the hottest page on Yahoo Finance, pulling rivals like AMD and Intel along for the ride.

The $11 billion guidance that broke Wall Street’s models

The real shocker was not just the backward-looking numbers, but Nvidia's astonishing forecast for the upcoming quarter.

Nvidia projected second-quarter revenue of roughly $11 billion, give or take 2%. Wall Street analysts had penciled in just $7.2 billion. That massive $3.8 billion delta sent analysts scrambling to trash and rewrite their financial models overnight.

Founder and CEO Jensen Huang explained on the call that the surge reflects a foundational transition from general-purpose CPUs to accelerated computing across global data centers.

Huang noted that the industry is at the start of what he expects to be a 10-year transition to essentially rebuild the world’s data center infrastructure around GPUs, smart switches, and SmartNICs, with generative AI dominating workloads.

Wall Street reacts: Scarcity, pure leverage, and few alternatives

Analysts across Wall Street quickly abandoned their conservative stances.

Wedbush analyst Dan Ives called the update a guidance for the ages, noting that while skeptics had been warning of an AI bubble, Nvidia demonstrated that it is the undisputed hearts and lungs of the AI revolution.

Citi analyst Atif Malik highlighted that out of the roughly $1 trillion worth of traditional data center CPUs installed over the last four years, only about 4% have been GPU-accelerated. In other words, the enterprise upgrade cycle is barely in its opening innings.

Baird analyst Tristan Gerra upgraded the stock with a $475 price target, pointing to overwhelming March channel demand for H100 chips sparked by ChatGPT and estimating that $10 in annualized earnings per share could be reachable within a couple of quarters.

Even cautious voices acknowledged the reality. Deutsche Bank raised its price target from $170 to $390, admitting that while data center digestion periods might loom down the road, Nvidia’s unmatched market leadership and the complete lack of viable alternatives make it nearly impossible to derail the current momentum.

Why it matters

In our view, Nvidia’s unprecedented quarter marks the moment generative AI shifted from software novelty to tangible, capital-intensive infrastructure reality.

For developers and startups, the immediate takeaway is clear: compute scarcity is going to stick around. As hyperscalers divert massive capex toward securing H100 clusters, renting cutting-edge GPU capacity will remain expensive and supply-constrained. Trying to train frontier foundation models from scratch is increasingly prohibitive; building specialized application layers on top of existing platforms looks far more sensible.

For the broader tech ecosystem, this signals a radical reshuffling of enterprise IT budgets. Dollars historically reserved for standard servers and CPU upgrades are now being cannibalized to fund GPU clusters.

Nvidia’s genius was spending years laying the proprietary software groundwork with CUDA and networking gear, leaving rivals virtually unable to catch up when the demand wave hit. The only real risk is whether cloud titans are currently panic-buying ahead of actual consumer utilization. When that initial buildout completes, Nvidia will face the challenging task of lapping these astronomical numbers.