1. Advanced cooling demand persists
- High-density clusters require advanced cooling, preserving differentiation and margins.
High-density training clusters sustain demand for advanced cooling, protecting JCI's differentiation and margins.
Inference commoditization pressures lower-spec cooling, capping JCI's addressable market and margins.
Bernstein remains constructive on the broader liquid-cooling opportunity, arguing that high-density training clusters will still require more advanced cooling architectures where technology differentiation, margins, and service attach rates remain strongest.
Google’s new Brazos liquid-to-air CDU, designed for lower-density AI inference deployments, could accelerate the standardization of simpler L2A cooling systems, potentially commoditizing portions of the CDU market.
Bernstein sees a risk that portions of the CDU market become more commoditized as inference workloads migrate into existing facilities, potentially pressuring pricing and margins for lower-spec cooling products.
Research through 2026-06-17 · 2026-06-17 to 2026-06-17 · 1 source reports.
| Range | Prices through |
|---|---|
| 1W | |
| 1M | |
| 3M | |
| 6M | |
| 1Y | |
| 2Y | |
| 3Y |
Research through 2026-11-03. Artifact validity does not establish factual accuracy or research freshness.