Summary

Zitron argues that NVIDIA's record revenues are almost entirely the product of speculative, years-in-advance GPU purchases by hyperscalers and neoclouds, the vast majority of which are sitting in warehouses or unpowered data centers rather than generating revenue. Using Microsoft as his central case study — which he says spent $265 billion in capex to put only ~$50 billion worth of GPUs online — he estimates that $200 billion or more in AI chips across the industry remain uninstalled. He contends that hyperscalers have systematically obscured this reality by using vague 'gigawatt capacity' language that conflates AI-specific infrastructure with general data center and power-procurement announcements. The artificial demand signal created by OpenAI and Anthropic — which absorb 70%+ of available AI compute — masks the fact that real, diverse market demand is a fraction of what's needed to justify the buildout. Zitron concludes that an inevitable overbuild collapse is coming, one he believes will be worse than the dot-com bust.

Key Insight

NVIDIA's record revenues are a lagging indicator of speculative GPU pre-purchasing, not a real-time signal of AI compute demand — meaning the entire financial narrative justifying the AI infrastructure boom is built on a systematic misrepresentation of how much capacity is actually operational.

Spicy Quotes (click to share)

• 5

  Today's newsletter is about a cruel truth: that NVIDIA's revenue growth is almost entirely the result of speculative purchases by hyperscalers and neoclouds that take years to install its GPUs.

• 7

  It appears that Microsoft is being vague with its numbers to hide the very obvious truth: that it's spending hundreds of billions of dollars to buy GPUs that sit in warehouses or unpowered data centers, likely years in advance.

• 9

  Jensen Huang saying that 'tokens are profitable and compute is revenue' as the vast majority of his sales generate neither tokens nor revenue because the fucking data centers take so long to build.

• 7

  The vast majority of companies that have bought GPUs have yet to turn them into meaningful revenue, if they've turned them on at all. Most of NVIDIA's sales are sitting in warehouses, and that is a significant disclosure that NVIDIA should have already been forced to make.

• 8

  Nobody on God's green Earth is sitting here wondering if Microsoft added gigawatts of CPU capacity or cloud storage — they're asking where is all the money you're spending on AI going, and the answer, it appears, is 'into a warehouse' or 'into a data center without power.'

• 6

  Everybody who participated in this obfuscation owns part of what comes next.

• 9

  None of this ever had anything to do with AI, and everybody who cheered Jensen Huang's ascent in the belief it did is a mark.

Tone

investigative, alarmed, combative