Every dollar of model intelligence flows through silicon. AI compute is the gating constraint and the largest capex line in the trade. The companies designing, fabricating, and networking the chips that train and serve frontier models. The incumbent (NVIDIA) and the credible challengers (AMD, Broadcom custom silicon, Cerebras for inference) sit on top of a stack that runs back through TSMC and ASML. When the trade has been wrong, it's been wrong here first.
AI Compute · 1d avg
-1.57%
8 tickers tracked
Signals (three independent reads)
Fund momentum
0
Funds steady — no meaningful 13F changes this quarter.
Sentiment trend (7d)
-2
17 articles · 0% bullish · split.
Canon citations
4
3 bullish, 1 structural from Aschenbrenner, Sutton, Sequoia, Andreessen.
What the canon says
“The trillion-dollar cluster is coming. Compute is the binding constraint, not capital.”
Aschenbrenner · bullish
“Methods that leverage computation are ultimately the most effective, and by a large margin.”
Sutton · structural
“Compute is the picks and shovels. Demand from the application layer compounds.”
Sequoia · bullish
“The cost of intelligence is falling on a Moore's Law-like curve. Every layer above benefits.”
Andreessen · bullish
Parallel readings, not a composite score. Composite scoring lands in v3 once each signal validates against subsequent moves.
Texas Gov. Greg Abbott delivered one of the starkest warnings yet from a Republican to the AI industry, saying data center companies "dug their own grave" and deserve the backlash they're facing after failing to win community support.