← Back to ProductsImmersion-Cooled Rack System
VasRack
Built for AI training clusters, enterprise AI, and high-density GPU deployments. The VasRack delivers 80–126 kW per cabinet with single-module or scalable deployment options using BYD modular server form factor.
126kW
Max per Cabinet
<1.1
pPUE Rating
BYD
Modular Servers
∞
Scalable Clusters
Technical Specifications
Built for scale.
| Cooling Capacity per Cabinet | 80–126 kW |
| pPUE | < 1.1 |
| Server Form Factor | BYD Modular |
| Coolant Type | Dielectric Coolant |
| Deployment Mode | Single-module or Scalable Cluster |
| GPU Support | NVIDIA L20, RTX 40/50, A Series, H Series |
| Monitoring | Full Modbus-TCP Telemetry |
| Redundancy | Active + Standby Pumps, Dual Fans |
| Operating Temperature | -10°C to 45°C |
| Power Supply | 400V |
Key Features
Enterprise-grade performance.
Ultra-high-density GPU deployment per cabinet
pPUE below 1.1 for maximum energy efficiency
BYD modular server architecture for flexible configuration
Scalable from single cabinet to multi-rack clusters
Compatible with all major NVIDIA GPU families
Full telemetry and remote monitoring capabilities
Active + standby pump redundancy for reliability
Optimized for tropical climate deployment
Full System View
Ready to scale your AI infrastructure?
Contact our engineering team to discuss VasRack configurations tailored to your deployment requirements.
Request a Quote