RockstarMarkets
All news
Markets · Narrative··Updated 2d ago
Part of: AI Capex

Nvidia's Circular Investment Thesis Sparks Durability Fears

Investors are questioning whether the AI capex boom is self-sustaining or a circular investment that risks collapsing if demand for AI services fails to materialize. Nvidia's buildouts at CoreWeave, Iren, and Corning show heavy exposure to infrastructure that may overshoot fundamentals.

R
Rocky AI · RockstarMarkets desk
Synthesised from 8 wires · 52 mentions in the last 24h
Sentiment
+30
Momentum
65
Mentions · 24h
52
Articles · 24h
46
Affected sectors
Related markets

Key facts

  • Nvidia invested in CoreWeave, Iren, Corning; all tied to AI infrastructure buildout
  • CoreWeave CEO warns Nvidia must expand capacity or lose customers to AMD; arms race forming
  • Circular investment concern: Nvidia profits from chips sold to build data centers that generate more demand
  • IBM: most companies now staffing chief AI officer roles; signals AI capex budgets being approved
  • Enterprise AI adoption accelerating, but revenue lag vs. capex remains unproven at scale

What's happening

A growing cohort of market observers has begun questioning whether the AI infrastructure capex cycle is truly durable or merely a circular bet that collapses once the reality of AI ROI emerges. Nvidia has made strategic investments in companies like CoreWeave, Iren (data centers), and Corning (fiber optics) that are building capacity for AI training and inference. The concern is that these investments fit 'squarely into the circular investment theme' that has been driving fears around the market's durability. In other words, Nvidia profits from selling chips to build data centers; those data centers then generate demand for more chips. If demand for AI services (and willingness to pay for them) fails to materialize, the entire edifice could crumble.

CoreWeave CEO has warned that Nvidia must expand AI capacity or risk losing customers to AMD. This suggests that competition is intensifying and that the capex race may be becoming an arms race driven more by fear of being left behind than by rational ROI calculations. If AWS, Google Cloud, or Microsoft Azure begin reporting AI margin compression (lower returns per dollar of capex), the thesis could unwind quickly. The supercycle narrative in semiconductors relies on this capex continuing unabated through 2027 and beyond.

On the flip side, the circular investment thesis assumes that AI services (like OpenAI's GPT offerings, Anthropic's Claude, or Google's Gemini) will fail to generate sufficient revenue to justify the capex. However, early data suggests enterprise AI adoption is accelerating; IBM's report on chief AI officer hiring shows boards are staffing AI roles, implying budget allocations. The debate hinges on whether capex will remain ahead of revenue growth or eventually normalize.

The risk is a recalibration where capex expectations fall sharply once the first few mega-cap cloud providers signal a slowdown in data center buildout or report disappointing AI monetization. This could cascade into downgrades for Nvidia, semiconductor makers, and equipment suppliers. For now, the momentum narrative dominates, but skeptics are circling.

What to watch next

  • 01Nvidia Q2 guidance; any sign of cloud customer capex slowdown or delay
  • 02AWS, Google Cloud, Microsoft Azure earnings calls; AI margin trends and ROI discussions
  • 03Cerebras IPO pricing: tests whether pure-play AI chip makers can sustain valuations
Mention velocity · last 24 hours
Coverage from these sources
Previously on this story

Related coverage

More about $NVDA

Topic hub
AI Capex: Who's Spending, Who's Earning, and What's at Risk

Tracking AI infrastructure capex — hyperscaler spend, data center buildouts, memory demand and the margin compression risk.