Nvidia’s Blowout Quarter Suggests the AI Boom Is Nowhere Near Its Peak

Record revenues, sold-out next-gen chips, and a sweeping vision of three massive platform shifts point to a company still accelerating into the future of computing.

When Nvidia reports earnings now, it’s less a financial event than a pulse check on the entire AI economy. If the boom were fading, Nvidia—supplier of the industry’s most essential hardware—would be the first to feel the slowdown. Instead, the company delivered another quarter that looked less like a crest and more like a launchpad.

Nvidia posted $57.01 billion in revenue and $1.30 in diluted earnings per share, topping estimates of $54.9 billion and $1.26. Sales rose 62% year-over-year, an almost absurd number for a company of this scale. Its data-centre business, fuelled almost entirely by demand for AI compute, generated $51.2 billion, beating expectations yet again. Then came the part markets were waiting for: the forecast. Nvidia expects around $65 billion in revenue next quarter—four billion more than analysts had pencilled in. For a company that has spent the past several years rewriting what “record quarter” even means, this one still managed to raise the bar.

Jensen Huang, Nvidia’s founder and CEO, opened the earnings call not with apologies for inflated expectations, but with an argument that the expectations themselves are still too small. Huang described the moment as one defined by three massive, simultaneous platform shifts. First is the move from general-purpose computing to accelerated computing, where GPUs replace CPUs as the fundamental engine of modern data centres. Second is the shift from traditional machine learning to generative AI, which requires enormous, increasingly sophisticated GPU clusters to train and run foundation models. And third is the transition toward what Huang calls agentic and physical AI—systems that don’t just predict but act, informing everything from autonomous factories to robots to next-generation vehicles.

Nvidia, he argued, isn’t just participating in these transitions; it is enabling all of them. “Each will contribute to infrastructural wealth,” Huang said, positioning the company not as a beneficiary of hype, but as the scaffolding beneath a global rebuild of compute infrastructure.

If the numbers weren’t enough to make that case, Huang added a more visceral signal. “Blackwell sales are off the charts, and cloud GPUs are sold out,” he said, referring to Nvidia’s next-generation architecture—the successor to Hopper, and the chip that virtually every major AI lab, cloud provider, and emerging startup has been scrambling to secure. The idea that demand for these chips is accelerating, not cooling, gives Nvidia an enviable problem: it simply cannot make enough of them.

From Huang’s perspective, AI has entered what he calls a “virtuous cycle.” More startups, more foundation model makers, more countries investing in sovereign AI programs—each new initiative creates more demand for compute, which leads to more model development, which further expands the ecosystem. “AI is going everywhere, doing everything, all at once,” he said, in a tone that sounded less like hype and more like inevitability.

So if there is an AI bubble, it’s doing a strange thing: growing while producing record-setting revenue and sold-out inventory. Investors, who had been bracing for signs of exhaustion, found none. Instead, they got a picture of a company still at the center of a global infrastructure build-out that is only beginning to mature.

For now, Nvidia remains the clearest indicator of where AI is headed. And based on this quarter’s results, that direction is still unmistakably upward.