Overview
- In mid‑May Nvidia reported record fiscal Q1 revenue of $81.6 billion, with $75.2 billion from its data‑center business, and publicly introduced the Vera Rubin CPU/platform to enter a roughly $200 billion addressable market.
- The company projected about $20 billion in standalone CPU revenue this year and said the Vera platform aims to move Nvidia beyond GPUs toward a full‑stack AI infrastructure role.
- Investor scrutiny intensified after Michael Burry published a Substack critique arguing Nvidia’s data‑center sales are concentrated in a few hyperscalers and that large non‑cancelable foundry commitments—reported near $119 billion—create structural risk.
- Nvidia responded to analysts after the criticism by pushing back on points about stock‑based compensation and depreciation and by defending the durability of its revenue and customer agreements.
- Beyond the headline numbers, the story highlights a deeper tradeoff: Nvidia’s ecosystem strength and measurable AI revenue at cloud providers support its case, while heavy reliance on a handful of buyers and large TSMC commitments raise the risk of sharp swings if hyperscaler spending slows.