McAllen, TX

Kati 2 / MFE-A

Northeast of Harlingen, TX in Willacy County

Data center coming soon

McAllen, TX

300+
MW
TOTAL CRITICAL IT
210K
SF
M2
WHITESPACE

AI-Ready Capacity with a 300+ MW Expansion Roadmap

100+MW Phase 1 Critical IT Capacity
300+ MW Total CIT Planned
Behind the meter, ERCOT + Renewable-Powered Campus

Metrobloks and Soluna have formed a joint venture to develop Project Kati 2, an AI-ready data center campus in South Texas. The site delivers immediate access to power and a clear path to scale, combining renewable energy economics with the reliability of the ERCOT grid.

Strategically Positioned at the U.S.–Mexico Gateway

  • Located in South Texas, within 40 miles of McAllen
  • Access to major long-haul fiber routes and key U.S.–Mexico cross-border network assets
  • Alternative AI siting option beyond Central & West Texas
  • Outside historic flood plains

The location supports low-latency connectivity for U.S., cross-border, and LATAM-oriented enterprise and AI workloads.

‍South Texas Power Advantage for AI Deployments

Project Kati 2 benefits from secured, renewable-powered infrastructure designed to support high-density AI and HPC workloads at scale. The campus combines low energy costs through renewable power with the reliability and resiliency of the ERCOT grid.

Designed for customers facing power constraints in traditional data center markets, MFE-A enables faster deployment timelines without waiting years for interconnection approvals.

The Perfect Mix for AI deployment

Dual-Source Power with Grid Resiliency and Low-Cost Renewable Power

MFE-A connects simultaneously to the ERCOT grid and Las Majadas wind farm, securing 490MW of total supply. This 100% renewable PPA structure delivers exceptional reliability and some of the lowest energy costs in North America, while maintaining full grid flexibility and uptime resilience.

Built for High-Density AI and HPC

Engineered for AI, HPC, and dense inference workloads, each data hall supports densities of 200kW+ per rack and beyond. A single-story architecture enables 100% direct liquid-to-the-chip cooling, with Tier III redundancy, 400 W/SF density, and scalable modules that grow with your compute needs.

MFE-A is engineered for demanding AI and HPC environments, supporting modern cooling and power architectures required by GPU-dense deployments.

Design Highlights

  • Built to VR200 standards (NVIDIA reference design)
  • Designed for 200kW+ per rack
  • 100% direct liquid-to-the-chip cooling
  • 800 VDC IT voltage architecture
  • Modular blocks enabling phased expansion
  • Tier III-aligned design principles
  • Dual-source power with scalable redundancy

Campus Specs

Critical IT:
300+ MW
Campus Size:
100+acres
Data Halls:
TBD
Configuration:
Multiple Buildings, Single Story
Whitespace:
210,000 sf
Availability:
Tier III compliant. 99.99% data halls / 99.999% network core
Connectivity:
Carrier Neutral
Power Density:
400 W per SF
Cooling:
100% direct liquid-to-the-chip
Sustainability:
Renewable energy (wind)
UPS:
BESS solution providing site-level N+1 for up to 1 hour
Power config:
800 VDC architecture
Generators:
Diesel generators provisioned if required by customer
Mechanical:
N+1 at Electrical Lineup level / N+4 at HU level

Perfect location for an AI Campus

Low latency connectivity to 27M people
Proximity to major metros including Austin, Houston, San Antonio and Monterrey, Mexico
Electric rates way below the national average, supporting energy intensive operations
Attractive tax incentives

A Campus Designed to Grow with Demand

The initial 100+MW phase is the first step in a larger campus roadmap supporting 300+MW of total Critical IT capacity. The modular design allows customers to scale efficiently as compute needs expand, without relocating or redesigning infrastructure.

A Campus Designed to Grow with Demand

The initial 100+MW phase is the first step in a larger campus roadmap supporting 300+MW of total Critical IT capacity. The modular design allows customers to scale efficiently as compute needs expand, without relocating or redesigning infrastructure.

Secure AI Capacity

Speak with a Metrobloks colocation specialist to discuss AI, HPC, and high-density deployment options.