Nvidia’s disappointment signals new phase in AI markets

The AI era is just beginning, writes Nigel Green, but the spoils might just be shared a little more widely than first thought.

Nvidia’s latest quarterly update has confirmed what many in the investment community have quietly suspected: the era of “hyper growth” for the world’s most prominent AI chipmaker is giving way to a phase of more measured, though still impressive, expansion.

The company delivered earnings that again surpassed expectations, with revenue climbing to nearly $47 billion and adjusted earnings per share topping forecasts. Its core data centre division surged by more than 50% year-on-year, underlining the continuing strength of demand for AI infrastructure.

But despite the scale of these achievements, markets focused on what was missing: margins slipping, revenue growth projections moderating, and evidence of structural vulnerabilities that will be harder to ignore.

The reaction was swift. Nvidia shares retreated in after-hours trading, a sign that investors are reassessing the assumptions that have propelled the stock to become one of the most important drivers of US equity indices.

Markets had been pricing the company as if its meteoric rise could go on forever. That illusion is now fading.

The concentration of Nvidia’s revenue is a critical concern. Reports indicate that two major clients, believed to be Microsoft and Meta, together account for almost a third of total sales.

For any business, that kind of reliance on a small number of customers is a glaring weakness.

For a company underpinning a huge proportion of the S&P 500’s performance, it is a systemic risk. Should either of those clients pull back on spending, the repercussions would reverberate through global markets immediately.

Competition is another headwind. Advanced Micro Devices and Intel are accelerating their product pipelines, and cloud giants are investing heavily in proprietary chips to lessen dependence on Nvidia’s offerings. Meanwhile, Washington’s restrictions on the export of high-performance processors to China – one of the largest potential markets – are creating barriers that domestic Chinese rivals are eager to exploit, backed by strong government support. Margins, once near 80%, are now sliding as these pressures mount.

This does not mean demand for AI hardware is slowing. Far from it. The appetite for computing power, data infrastructure, and the energy to fuel it all continues to grow exponentially. Global energy use for data centres is projected to more than double by the end of this decade.

The broader AI story remains one of transformational growth. But Nvidia’s share of that pie will not be as dominant as it once was.

The paradox for investors is clear. Nvidia remains at the heart of the AI revolution, but the sheer weight it carries in equity indices introduces fragility. Technology companies now make up close to 30% of the S&P 500, with Nvidia, Apple, and Microsoft together accounting for over 15%. Last year alone, Nvidia was responsible for more than a quarter of the index’s total returns.

This level of concentration is unhealthy. It leaves the market overexposed to the fortunes of a handful of stocks and diminishes diversification benefits that indices are supposed to provide.

This is why the current earnings season matters far beyond the fortunes of one company. It marks a subtle but significant shift: from a single champion driving the AI narrative, to a more complex and diversified growth story across the sector.

The winners of tomorrow will be found in semiconductors, but also in cloud infrastructure, energy utilities, cybersecurity, and software platforms building real-world applications. The investment opportunity is no longer about backing one company to dominate; it is about capturing a structural transformation across the global economy.

For investors, that means adjusting expectations and strategies. Nvidia remains a formidable business and a vital part of the AI ecosystem, but its path ahead will be defined by competition, customer concentration risks, and geopolitical headwinds as much as by relentless demand.

Allocations that assume indefinite, outsized returns from a single name are increasingly hazardous.

AI will continue to reshape productivity, business models, and even entire industries. The scale of spending and innovation pouring into this space is unlike anything we have seen in recent decades.

But markets evolve, and leaders cannot hold unchallenged dominance forever. Nvidia’s latest numbers show that reality is setting in. High growth is still powerful growth – but it is no longer hyper.

The AI era is just beginning, but its spoils will be shared more widely. Investors who recognise this shift and diversify across the expanding ecosystem stand to benefit most in the years ahead.