Tonight everyone will watch one number from Alphabet ($GOOGL) - does capex move? That’s not the most important question on the call.. A flat $180-190B vs a raise to $195B (we think ~40% chance) is rounding error against their ~$460B cloud backlog. What matters is WHAT the spend is buying. If management signals more servers and compute, the AI infrastructure thesis for NVDA, MU, AVGO and VRT remains intact even on a flat capex guide. If the language shifts toward land, construction, or slower deployment, that’s the real read-through
View on X ↗