Total Power
24 MW
ODESSA AI COLOCATION
High-Density AI Colocation:
Odessa Site
Redundant Web Services (RWS) delivers dedicated, high-performance compute infrastructure for demanding AI and deep learning workloads. Positioned specifically to capture high-density enterprise requirements and critical Oil & Gas AI telemetry workloads, our South and West Odessa sites offer the scalable capacity you need today.
24 MW
Total capacity
90d
Deploy time ARO
75%
vs hyperscaler cost
Accelerating the NVIDIA AI Infrastructure Frontier
HIGHLIGHTS
Facility & Architecture Highlights
-
Deployment Velocity
90 to 120-day lead time ARO for custom colocation build-outs.
-
Network Readiness
Sites are ready to populate with lite & dark fiber available immediately.
-
Next-Gen Hardware Support
Fully equipped to host the complete NVIDIA portfolio, including A100, H100, H200, B200, B300, and GB300 systems.
-
Container-Scale Deployment
Built on highly efficient 120 kW modules that are fully stackable. Our architecture is designed to seamlessly handle hyperscale rollouts of 50 to 200 containers per site.
-
Advanced Cooling
Each container module is adaptable and equipped to support both air and water cooling configurations.
-
Geographic Advantage
Proximity to the Permian Basin makes these sites a perfect geographic fit for energy companies and time-sensitive telemetry workloads.
ODESSA BENEFITS
The Odessa Advantage
Odessa combines utility-scale AI capacity with Permian Basin proximity, rapid deployment paths, and infrastructure designed for dense GPU workloads from day one.
A100 Capacity
14,400
H200 Capacity
13,200
B200 Capacity
12,000
SITE SPECIFICATIONS
Other Site Capacities & GPU Densities
Our infrastructure allows you to match infrastructure to workload without compromising performance or efficiency.
| Site Location | Total Power | A100 Capacity | H200 Capacity | B200 Capacity | H100, B300, GB300 |
|---|---|---|---|---|---|
| George West, TX | 13 MW | 7,800 | 7,150 | 6,500 | Supported |
| MacAllen, TX | 10 MW | 6,000 | 5,500 | 5,000 | Supported |
| Wenatchee #1, WA | 6 MW | 3,600 | 3,300 | 3,000 | Supported |
GPU capacities are based on standard per-MW densities inside our 120 kW modules: A100 (600/MW), H200 (550/MW), and B200 (500/MW).
WHY RWS
The Advantage of Partnering With RWS
Enterprise Reliability without the Drag
Skip the 12 to 18-month industry standard hyperscaler wait times and deploy faster.
Predictable Economics
Transparent, reserved pricing resulting in 40–75% cost savings compared to hyperscaler on-demand pricing.
Architectural Freedom
We support tenant-supplied hardware, hybrid colocation, or RWS-provided hardware with zero hyperscaler vendor lock-in.
Contact Us to Reserve Capacity
To discuss site tours, technical validation, or to reserve your colocation deployment, please contact: