Artificial Intelligence
Noviant Launches LiquidEdge AI POD: Liquid-Cooled AI Datacenter Container Solution
May 14, 2026
Noviant embraces the AI era by introducing the LiquidEdge AI POD — a factory-built, plug-and-play liquid-cooled AI datacenter container deployable in just 8 weeks, supporting NVIDIA GB200, H100, B200, RTX 5090, and AMD MI300X GPUs.
LiquidEdge AI POD — 40-ft Liquid-Cooled AI Datacenter Container
As artificial intelligence reshapes industries worldwide, Noviant is proud to announce our partnership with LiquidEdge to bring the AI POD liquid-cooled datacenter container solution to enterprise clients across New York and North America. This marks a significant step in Noviant's commitment to delivering cutting-edge AI infrastructure that meets the demands of the modern intelligence era.
The global data center industry faces an unprecedented challenge: AI compute demand is growing exponentially while traditional datacenter supply lags years behind. Conventional builds still require 36+ months due to grid interconnect queues, permitting delays, and construction lead times. The LiquidEdge AI POD was engineered specifically to solve this bottleneck.
The LiquidEdge AI POD is a 40-ft North American-spec ISO container housing a complete, factory-assembled liquid-cooled AI datacenter. Built and tested at the factory, it arrives on-site ready for connection — no civil works required, with fewer than 5 field crew members needed at installation.
- Factory-built and pre-tested before shipment
- Direct liquid cooling (DLC) achieving PUE ≤ 1.15
- Operating range: -40°C to +50°C, IP55 outdoor-rated enclosure
- 8-week factory-to-live delivery timeline
- Fully relocatable and redeployable
- North America compliant: UL 891, NEC 2023, NFPA 75/76, ASHRAE TC 9.9
The LiquidEdge AI POD is available in three configurations to match different workload requirements:
LE-44 (Single-Rack / Edge): 44 kW IT load, supporting up to 64× NVIDIA RTX 5090, 32× H100/H200, or 24× B200 GPUs. Ideal for AI inference, fine-tuning, and edge AI deployments.
LE-88 (Twin-Rack / Mainstream): 88 kW IT load, supporting up to 128× RTX 5090 or 64× H100/H200. Designed for Neocloud operators and GPU-as-a-Service providers.
LE-200 (Quad-Rack / Heavy): 200 kW IT load, supporting one complete NVIDIA GB200 NVL72 system (72 GPUs) or up to 256× RTX 5090. Built for large language model training and hyperscaler bridge capacity.
All three configurations use direct liquid cooling (DLC) with cold-plate technology that captures ≥80% of GPU heat into the coolant loop, enabling the industry-leading PUE ≤ 1.15 — dramatically better than the 1.50–1.70 PUE typical of air-cooled facilities.
The AI POD also features an integrated AIOps platform with 4,000+ telemetry points per unit, a CFD-validated digital twin for workload simulation, and predictive maintenance capabilities that flag hardware degradation 72+ hours in advance.
Noviant is now accepting inquiries and pre-orders for the LiquidEdge AI POD. Our team will work with you to assess your power availability, GPU requirements, and deployment timeline to identify the right configuration for your business. Contact our AI infrastructure team today to schedule a consultation.