A fresh $1.5 billion capital-expenditure plan will fund Meta’s new AI-optimized datacenter in Texas, intended to support training and inference workloads across future product roadmaps. Rather than incremental upgrades, the facility is designed around high-density compute, advanced liquid cooling, and renewable power tie-ins. Management targets full operations by 2026, citing expected efficiency gains for recommendation engines and generative features. Analysts view the move as a strategic capacity build supporting revenue growth, though near-term operating expenses and depreciation are likely to rise as construction advances.