Meta announced a new AI-optimized data center campus in El Paso, Texas, expanding the company’s infrastructure footprint to support next-generation AI workloads across its apps and services. The campus is designed from the ground up for high-density compute used in model training and inference, pairing modern network fabrics with advanced thermal management and liquid cooling to run AI systems more efficiently. Meta says the facility aligns with its long-term roadmap for custom silicon and heterogeneous compute, enabling flexible deployment of GPUs alongside internal accelerators as models evolve.

The El Paso design continues Meta’s sustainability approach: clean-energy procurement, water-efficient cooling strategies, and building standards intended to minimize embodied carbon over the campus lifecycle. The company notes that siting in West Texas provides access to strong renewables growth and long-haul network routes, improving resiliency and latency for users and developers who rely on Meta’s AI features. Beyond core training clusters, the campus will host scaled inference services that power everyday AI assistants, ranking systems, and generative tools across Facebook, Instagram, WhatsApp, and Reality Labs devices.

For the region, Meta highlights opportunities for local suppliers, construction trades, and ongoing data-center operations roles once the site is live. The announcement underscores how major AI platforms are racing to add specialized capacity: facilities must handle dense racks, liquid cooling, and rapid refresh cycles while staying within strict energy and sustainability targets. By bringing an AI-first blueprint to El Paso, Meta aims to create a flexible campus that can accommodate future model classes without disruptive rebuilds.

Read Meta’s announcement here: Meta’s new AI-optimized data center in El Paso.