The artificial intelligence boom continues to drive enormous investment across the global technology industry, but analysts are increasingly divided over how long the current pace of spending can continue. Nvidia’s latest results have provided a major boost to the bullish case, with quarterly revenue reaching $96.2 billion, up 106% year over year, while its data-centre business generated $89 billion. Nvidia also forecast about $108 billion in revenue for the current quarter.
The numbers have temporarily eased fears that AI investment is approaching a peak. Nvidia shares jumped 8.7% after the results, while the Nasdaq gained 1.57% and the S&P 500 technology sector rose 3.4%. However, analysts say the bigger question is no longer whether companies are willing to spend on AI, but whether that spending will eventually generate enough revenue and profits to justify the enormous investment.
Strong Nvidia results strengthen the bullish case
Nvidia has become one of the most important indicators of the health of the AI economy because its chips power many of the world’s largest AI data centres.
The company’s latest results showed that demand remains exceptionally strong. Data-centre revenue more than doubled from a year earlier, while Nvidia expects another substantial increase in revenue in the coming quarter. The company is also forecasting around 70% revenue growth for fiscal 2028, a figure that was considerably above some market expectations.
For bullish analysts, these numbers suggest that AI spending is not simply speculative enthusiasm. Major technology companies are continuing to spend billions of dollars building computing capacity because they expect AI services to become a significant source of future revenue.
Amazon, Google, Microsoft and Meta are among the companies investing heavily in AI infrastructure, creating enormous demand for GPUs, networking equipment, memory and data-centre capacity.
But analysts are asking a much bigger question
Despite the strong results, concerns about sustainability have not disappeared.
The biggest issue is the return on investment from AI infrastructure. Technology companies are committing hundreds of billions of dollars to data centres and AI chips, but investors ultimately want evidence that these investments will generate sufficient profits.
Reuters recently highlighted concerns that the AI earnings boom could face a “data-centre reality check” if companies fail to generate adequate returns from their enormous infrastructure spending.
This creates a critical test for the industry. AI companies can continue purchasing more GPUs and building more data centres, but eventually investors will want to see strong revenue growth from the AI products being developed with that infrastructure.
The spending cycle is still accelerating
For now, however, there is little evidence that the major technology companies are preparing to slow their investment.
Google, for example, raised its 2026 capital expenditure forecast to between $195 billion and $205 billion, while Google Cloud revenue increased 82% in the second quarter.
That is important because the sustainability of the AI boom depends heavily on spending by the large cloud companies. If Amazon, Microsoft, Google and Meta continue increasing their capital expenditure, chipmakers such as Nvidia, AMD, Broadcom and memory manufacturers could continue benefiting.
The problem would arise if these companies eventually conclude that additional AI infrastructure is producing diminishing returns.
AI chip demand remains exceptionally strong
Another argument supporting the sustainability of the boom is the continuing demand for advanced semiconductors.
South Korea’s semiconductor industry is benefiting heavily from AI demand, with Samsung Electronics and SK Hynix seeing strong demand for advanced memory used in AI systems. A Reuters poll expects South Korean exports to remain extremely strong in August, largely because of semiconductor demand.
TSMC has also raised its revenue outlook and increased capital spending to meet demand for AI chips.
These developments suggest that the AI investment cycle has spread well beyond Nvidia. The broader supply chain—from chip designers and foundries to memory producers and data-centre infrastructure companies—is benefiting.
Investors are becoming more demanding
At the same time, investors are no longer rewarding every company simply for announcing an AI strategy.
AMD recently saw its shares fall after investors demanded stronger evidence that its AI business could deliver greater returns.
Marvell Technology provides another example. The company raised its revenue forecasts because of growing demand for custom AI chips, yet its shares fell after investors questioned how quickly some major AI deals would translate into revenue.
This indicates an important change in market behaviour. Investors still believe in AI, but they increasingly want actual earnings, cash flow and measurable returns, rather than promises about future AI growth.
The biggest risk could come from AI customers
One of the less obvious risks is that Nvidia’s customers are themselves spending enormous amounts of money on AI infrastructure.
Companies such as Microsoft, Amazon, Google and Meta can afford these investments because of their enormous existing businesses. But the question is whether AI-related revenue will eventually grow quickly enough to justify the spending.
If AI applications become highly profitable, the current infrastructure boom could continue for years.
If AI products fail to generate sufficient returns, companies could eventually reduce capital expenditure. That would have a major impact on Nvidia and the entire semiconductor industry.
This is why analysts increasingly view AI monetisation as the next major test.
Nvidia’s margins reveal another challenge
Nvidia itself has warned about rising costs.
The company expects gross margins to decline toward approximately 72%-73% next fiscal year, partly because of higher memory costs.
Although that remains an extremely strong margin, the decline shows that maintaining rapid growth is becoming more complicated.
Nvidia is also facing increasing competition from AMD, Intel and custom AI chips developed internally by major technology companies such as Google and Amazon.
If customers increasingly develop their own processors, Nvidia could eventually face pressure on market share and pricing.
Concerns over valuations and financial structures remain
Some analysts are also concerned about the enormous valuations attached to AI-related companies.
Nvidia’s market capitalization has reached roughly $5.4 trillion, illustrating how much investor expectations are already built into the sector.
There are also concerns about financing arrangements within the AI ecosystem. Nvidia has provided substantial financial support to some AI infrastructure partners, raising questions among investors about whether some demand could be indirectly supported by the companies supplying the technology themselves. Reuters reported that Nvidia recently paused a revenue-sharing financing initiative aimed at smaller AI cloud companies amid investor concerns about such arrangements.
These concerns do not necessarily mean the AI boom is a bubble, but they demonstrate why analysts are paying closer attention to the quality of AI-related growth.
Analysts remain broadly optimistic—but more cautious
The latest Nvidia results have shifted the debate. Instead of asking whether the AI boom is about to collapse, investors are increasingly asking how long the current growth rate can continue.
Some analysts remain extremely bullish. Nvidia’s latest outlook has led analysts to raise their expectations, with some arguing that the company could eventually approach $1 trillion in annual revenue if AI infrastructure demand remains strong.
Others remain cautious, warning that the enormous spending cycle must eventually be supported by real economic returns.
The market’s reaction suggests that, for now, the bullish camp has the advantage. Nvidia’s results have shown that demand remains strong, while major technology companies continue to increase AI investment.
The next phase will be about returns
The AI boom therefore appears to be entering a new stage.
The first phase was about building AI infrastructure. Companies rushed to acquire GPUs, construct data centres and develop increasingly powerful models.
The next phase will be about monetisation.
Investors will want to know whether AI-powered search, software, cloud services, advertising, automation and other applications can generate enough revenue to justify the enormous capital being deployed today.
For now, Nvidia’s results provide powerful evidence that the AI boom remains alive and that demand for computing infrastructure is still accelerating. But analysts are increasingly focused on what happens next.
The central question is no longer whether the world will spend more on AI—it is whether the returns from that spending will be large enough to sustain the boom. If AI companies can demonstrate strong and recurring profits, today’s investment cycle could continue for years. If returns disappoint, the same enormous spending that has driven the technology rally could eventually become its biggest vulnerability.
About The Author
