Nvidia Earnings Smash Expectations – and Prove AI Bubble Fears Wrong

Nvidia Earnings Smash Expectations – and Prove AI Bubble Fears Wrong

For the past several weeks, fears about an ‘AI bubble’ have been hammering AI stocks

Big Tech’s soaring capex has traders worried that firms are overspending only to underdeliver – and that’s been weighing heavily on the AI trade. 

This is illustrated by the Global X Artificial Intelligence & Technology ETF (AIQ) – a strong proxy for the industry – which has been on a downward slope since late October.

That slump mirrored fading confidence in AI demand – until Nvidia (NVDA) changed the tone last night… and resuscitated the AI stock rally

The tech titan delivered a monster third-quarter earnings report – with even more monstrous guidance for Q4; the sum of which confirms that the AI frenzy is speeding up, not slowing down. 

Revenues rose 62% in Q3 and are expected to rise 65% next quarter. This is the first time since late 2023 that Nvidia’s sales growth rate has re-accelerated in two consecutive quarters.

After six straight quarters of slowing growth, the AI plumbing just hit the gas again – and the numbers tell the story…

Nvidia Earnings Overview: The AI Engine Is Accelerating Again

Let’s go over a quick rundown of Nvidia’s latest numbers so you can see just why we’re so bullish here:

  • Q3 revenue: $57 billion, up 62% year-over-year (YoY) and 22% quarter-over-quarter (QoQ)
  • Data center revenue: $51.2 billion, up 66% YoY and $10 billion sequentially
  • Q4 guidance: ~$65 billion in revenue, up ~65% YoY at the midpoint
  • Gross margin: ~73-75% non-GAAP – at this scale

We feel it’s important to highlight that this isn’t some tiny high-growth software start-up. Nvidia is running at a $200 billion-plus annualized revenue pace and still compounding north of 60%.

And growth is picking up steam. 

For the past six quarters, Nvidia’s sales growth rate had been slowing. Bears believed it to be “the beginning of the end.” But what this quarterly performance just proved is that AI demand wasn’t dying – it was digesting before its next leg up.

Now:

  • Growth has re-accelerated to 62%.
  • Management is calling for even faster growth next quarter at 65%.
  • And that’s with essentially zero China data center revenue baked in

If you’re looking for evidence that the AI Boom is “slowing,” you won’t find it in Nvidia’s financials.

And that’s just from what we can see on the surface… 

The AI Boom Is Speeding Up, Not Slowing Down

Under the hood, the story is even more bullish than what the headline numbers suggest.

Cloud Providers Are Sold Out

On the Nvidia’s conference call, management said the clouds are sold out and that the GPU installed base – Blackwell, Hopper, and older Ampere – is fully utilized

Translation: there is no sign of hyperscalers slamming on the brakes. If anything, they’re still flooring it.

With current GPU capacity tapped, attention now turns to how far hyperscalers are willing to expand.

AI Capex Spending Is Surging

That full utilization is exactly why Nvidia now sees visibility into roughly $500 billion of Blackwell + Rubin revenue through the end of 2026 – and that number has been increasing as new AI factory deals get signed.

Meanwhile, external estimates now see AI infrastructure spending heading toward $3- to $4 trillion by 2030

If this were a bubble that was about to pop, we’d be seeing missed estimates, weak guidance, slowing orders, and contracting capex plans.

Instead, we’re witnessing beat-and-raise performance, re-accelerating growth, and bigger long-term capex envelopes.

The ‘AI Bubble’ Fear Just Got Debunked

A recent fund manager survey from Bank of America (BAC) said the ‘AI bubble’ is the No. 1 perceived tail risk, with ~45% of managers citing it. 

In effect, Nvidia’s response to Wall Street’s top fear was: ‘Appreciate the concern. See: 62% growth, 65% forward guidance, and sold-out clouds.’

Jensen Huang literally said he doesn’t see an AI bubble – he sees real, widespread demand across clouds, enterprises, and emerging “agentic” and physical AI use cases. And Nvidia’s numbers back him up.

admin