CoreWeave guided full-year 2026 revenue to $12.4 billion to $13.2 billion after the bell Tuesday, with $960 million to $1.15 billion in adjusted operating income. The stock climbed 8% in after-hours trading. Supermicro also beat expectations. The results arrived hours after the Nasdaq closed down 0.6%, dragged by Amazon, Apple, Google, and Microsoft.
What the Numbers Actually Say
CoreWeave went public earlier this year. It is already guiding to revenue that most S&P 500 companies never reach. The midpoint of the guidance ($12.8 billion) would make it one of the fastest-scaling infrastructure companies in recorded history. Analysts surveyed by LSEG had expected $12.63 billion in full-year revenue. CoreWeave came in above.
The company provides GPU computing that powers AI training and inference. Its customers include OpenAI and Microsoft. Revenue growth at this pace confirms that AI workloads are not merely a capital expenditure story for the hyperscalers. CoreWeave is a separate layer in the stack, and that layer is filling up faster than the market expected.
Supermicro's beat reinforces the same point from the hardware side. SMCI builds the server infrastructure that sits between the GPU and the data center. Two AI infrastructure companies, two beats, same night.
The Aschenbrenner Context
The timing is clarifying. Leopold Aschenbrenner's Situational Awareness fund ran 4x leverage on the AI infrastructure thesis and was forced to unwind most of its positions last week after CoreWeave and related names sold off. The fund peaked at nearly $45 billion in assets.
The thesis was correct. CoreWeave at $12 billion in annual revenue proves that AI compute demand is structural and accelerating. The fund's mistake was not the thesis. It was applying 4x leverage to a volatile position in a company that had just gone public and was still finding its market valuation.
Aschenbrunner is already back raising fresh capital. The AI thesis was not broken by the blowup. The fund was broken by the leverage.
The Bearish Signal Inside the Bull Print
CoreWeave's revenue depends almost entirely on a small number of large customers. OpenAI and Microsoft represent concentrated counterparty risk. If either customer reduces cloud spending, CoreWeave's guidance becomes optimistic. The company is also still loss-making on a GAAP basis, narrowing losses rather than turning profitable.
The broader market is asking whether AI capex can continue at current rates. Amazon, Apple, Google, and Microsoft dragging the Nasdaq down on the same day CoreWeave beat is not an accident. The market is skeptical of Big Tech AI returns even as it rewards pure-play AI infrastructure.
The Infrastructure Layer Is Real
Bitcoin miners occupy an adjacent position. They compete for the same GPU inventory and power infrastructure as AI data centers. Aschenbrenner's forced selling hit MARA, RIOT, and CLSK regardless of their fundamentals. CoreWeave's earnings restore the credibility of the thesis that Bitcoin mining and AI compute share a structural demand floor.
The infrastructure layer beneath AI is real, growing faster than expected, and generating real revenue. The speculation was in the leverage, not the sector.
Follow @UTXOMacro for daily Bitcoin, macro, and AI breakdowns.