Capex Flow Map
Two-stage Sankey of AI-attributable hyperscaler capex routed through infrastructure layers to specific suppliers. The sensitivity panel below lets you stress-test the layer-mix assumption live, useful when an analyst pushes back on the default baseline. Every number is a modeled estimate: see Methodology for the math and assumptions.
Total AI-attributable capex in scope: $370.4B
Stress-test the model. Adjust how much of each hyperscaler dollar lands in each infrastructure layer; the Sankey and supplier table re-rank instantly. Values normalize to 100.0% so the underlying flows stay mass-conserving.
Re-ranked as the layer mix changes. Compare against the default mix to see which names are sensitive to which assumption.
| Supplier | Primary layer | Modeled $B | Share |
|---|---|---|---|
| NVDA NVIDIA | Compute Silicon | $122.1B | 33.0% |
| AVGO Broadcom | Compute Silicon | $39.8B | 10.7% |
| MU Micron Technology | HBM & Memory | $37.0B | 10.0% |
| VRT Vertiv Holdings | Power & Electrical | $36.1B | 9.8% |
| ETN Eaton | Power & Electrical | $25.6B | 6.9% |
| PWR Quanta Services | Datacenter Build | $24.4B | 6.6% |
| ANET Arista Networks | Networking & Optical | $16.3B | 4.4% |
| GEV GE Vernova | Power & Electrical | $13.9B | 3.8% |
| AMD Advanced Micro Devices | Compute Silicon | $13.7B | 3.7% |
| TSM TSMC | Foundry & Equipment | $11.7B | 3.2% |
| CRWV CoreWeave | Neoclouds | $10.2B | 2.7% |
| ASML ASML Holding | Foundry & Equipment | $5.2B | 1.4% |
| AMAT Applied Materials | Foundry & Equipment | $5.2B | 1.4% |
| CIEN Ciena | Networking & Optical | $4.4B | 1.2% |
| LRCX Lam Research | Foundry & Equipment | $3.9B | 1.1% |
It is not a revenue model. Hyperscaler capex is the spend; supplier revenue depends on shipment timing, mix, customer concentration, and competitive substitution. The flow map gives an upper-bound proxy for the addressable dollar base.
Internal accelerators are partially excluded. Google TPU is captured via AVGO (designer); Amazon Trainium and Microsoft MAIA are not in the supplier roster, so the "sum of shares" for AMZN and MSFT compute_silicon intentionally is <100%: the residual stays internal.
HBM is partial. Micron is the only HBM name in scope; SK Hynix and Samsung dominate. The shown $B for memory understates industry-wide HBM TAM by ~3×.