Nvidia’s blockbuster quarter resets the AI narrative

NVIDIA’s latest result has landed at a sensitive point for global equities, after a sharp tech correction over the past month rattled confidence in the AI trade and had traders wondering whether the theme was running out of steam. Rather than confirming the bears’ fears, the company has delivered a blockbuster print that resets the tone and reaffirms the core thesis that AI infrastructure spending is no fad.

Third-quarter FY26 revenue came in at a record US$57.0 billion versus expectations closer to US$55 billion, up 22 per cent on the previous quarter and 62 per cent year on year, with data-centre revenue surging to US$51.2 billion. Earnings per share of US$1.30 comfortably beat consensus around US$1.25, gross margins held at an extraordinary 73–74 per cent, and management guided to roughly US$65 billion of revenue in the current quarter.

The market reaction was swift: NVIDIA shares jumped more than 5 per cent in after-hours trade as investors digested not just the beat, but the message embedded in the numbers. This is a business that has become the de facto utility provider of the AI age. Data-centre sales – the heart of NVIDIA’s AI franchise – grew 66 per cent year on year, with around US$43 billion coming from “compute” (GPUs and systems) and US$8.2 billion from networking gear that allows tens of thousands of chips to act as a single computer. CEO Jensen Huang’s verdict was blunt: Blackwell sales are “off the charts”, cloud GPUs are “sold out”, and the industry has entered a “virtuous cycle of AI” where demand for training and inference is compounding simultaneously.

What makes this quarter more than just another beat is the broader context. In recent weeks, heavyweight investors such as SoftBank and Peter Thiel have trimmed NVIDIA holdings, feeding a narrative that AI might be the next crowded bubble. At the same time, hyperscalers including Microsoft, Amazon, Alphabet and Meta have all lifted capex guidance, collectively pointing to more than US$380 billion of annual spend, much of it directed to AI infrastructure. NVIDIA now adds another piece to that puzzle: visibility on roughly US$500 billion of orders across 2025–26, and a finance chief who openly argues the company is positioned as “the superior choice” for what could become US$3–4 trillion a year in global AI infrastructure build-out by the end of the decade.

That combination helps answer the question on every CIO’s and portfolio manager’s mind: is this a speculative mania, or the early innings of a genuine industrial shift? There is no doubt that parts of the market have run ahead of themselves and that corrections will remain a feature, not a bug, of an AI-driven tape. But it is hard to call it a pure bubble when the world’s largest cloud and internet platforms are locking in multi-year supply, governments are competing to host “AI factories”, and customers from telcos to carmakers are embedding NVIDIA roadmaps into their own strategic plans. In that sense, recent weakness looks more like a valuation reset inside a powerful structural story than the end of the trade.

For industry and global trade, NVIDIA’s result is a lighthouse signal. Upstream, it entrenches the importance of advanced manufacturing hubs – from TSMC’s fabs in Taiwan and Arizona to memory suppliers in Korea and networking specialists across the US, Israel and Europe. Downstream, it accelerates the pressure on enterprises, including in Australia, to modernise data centres, upgrade networks and rethink energy supply to support ever-denser compute. As AI workloads scale, so too will cross-border flows of chips, capital and data, tightening the links between digital infrastructure policy and traditional trade and investment strategies. Countries that can offer stable power, clear regulation and access to skilled engineers will find themselves competing to anchor this new layer of critical infrastructure. 

Zooming out, NVIDIA is no longer just another chip designer; it is increasingly the architecture layer for what many now describe as the fourth industrial revolution. Its platforms underpin everything from foundation model training to industrial robotics, autonomous vehicles and “digital twins” of factories and cities. Huang’s rhetoric about AI “going everywhere, doing everything, all at once” may sound grandiose, but the financials now arriving each quarter give that line real weight. For boards, policymakers and investors, the message is clear: this is not merely a story about one company’s margin profile. It is about whether economies are prepared to plug into – and help govern – an AI infrastructure stack that will shape productivity, competitiveness and even security over the next decade. On this evidence, the AI build-out is very real, and NVIDIA remains the company the rest of the world is forced to benchmark against.

Get the stories that matter to you.
Subscribe to Cyber News Centre and update your preferences to follow our Daily 4min Cyber Update, Innovative AI Startups, The AI Diplomat series, or the main Cyber News Centre newsletter — featuring in-depth analysis on major cyber incidents, tech breakthroughs, global policy, and AI developments.