What Is vCloud?
Hyperscale AI compute with cloud‑level agility and sovereign‑grade control for enterprises, governments, and AI innovators.
HubOne vCloud is a next‑generation AI infrastructure platform that combines hyperscale data center capacity with a programmable cloud control plane. It is designed for organizations that need massive GPU compute, deterministic performance, and strict data sovereignty without giving up the flexibility of modern cloud services.
Built on a 160,000 sq ft hyperscale campus, HubOne vCloud turns traditional static data center infrastructure into a programmable AI fabric. Compute, power, and network resources can be dynamically allocated across enterprise, government, and AI developer workloads from a single control plane.
Resilient, Redundant, Reliable. Simple.
Contracts available now for 2027.
Hyperscale Campus Architecture
Sovereign‑Grade Hyperscale Design
The HubOne vCloud Campus
World-class engineering mesh of four hyperscale facilities,
together delivering capacity for approximately 2,000 AI racks
within 160,000 sq ft of programmable infrastructure.
Key Infrastructure Features:
8 fractional compute zones (≈20,000 sq ft each)
10–15 MW of IT load per fractional block
Liquid‑to‑chip cooling for ultra‑dense GPU clusters
Support for NVIDIA and AMD GPU architectures
High‑capacity optical networking fabric across the campus
This modular architecture allows incremental deployment, rapid scaling, and flexible tenant allocation, without requiring a full campus build‑out at launch.
From Static Data Center to Programmable AI Platform
A New Model for AI Infrastructure
Traditional hyperscale data centers operate as fixed assets that are slow to reconfigure and hard to program. HubOne vCloud introduces a programmable infrastructure model, where compute, networking, and power can be provisioned and reallocated in software.
With HubOne vCloud, you gain:
Fractional hyperscale deployment starting at 20,000 sq ft blocks
Automated scaling across the full 160,000 sq ft campus
On‑demand provisioning of AI training and inference capacity
Dedicated sovereign compute zones for regulated workloads
SLA‑backed AI infrastructure services operated as vCloud offerings
The result is a hybrid between hyperscale physical infrastructure and sovereign cloud: physically grounded, but cloud‑native in how it is operated and consumed.
where we operate
Orchestration + NXS Infrastructure
Physical Infrastructure Layer – NXS Hyperscale Facilities at North Gate sites in Alberta, Canada.
The NXS campus provides the underlying AI‑optimized infrastructure layer for HubOne:
- High‑density GPU clusters for training‑ and inference‑grade workloads
- Deterministic power chain (HVDC → UPS → BESS → rack distribution)
- Liquid‑to‑chip cooling for advanced, high‑TDP AI hardware
- Carrier‑grade optical connectivity and an integrated optical fabric
- 20,000 sq ft blocks functioning as fractional hyperscale units
Coverage includes Canada, one third of the United States, and parts of Europe, please call for details.
The Campus Operating System
Digital Control Layer
On top of the physical infrastructure of HubOne, NXS operates vCloud and runs a digital control layer, the operating system of the entire campus:
- Digital twin inventory for power, racks, cooling, and connectivity
- Automated service orchestration across compute and network resources
- AI‑driven assurance and AI Ops monitoring
- Closed‑loop automation to maintain SLA and compliance targets
Together, these layers create a single operational control plane for managing hyperscale AI infrastructure as a programmable, policy‑driven system - built for performance.
AI as a Service
Service Layer Example Service Tiers:
- Fractional NXS AI Training Tier
- Fractional NXS Inference Tier
- Dedicated 400G optical connectivity slices
- Sovereign AI secure zones
- Private AI cluster deployments
End‑to‑End Automated Provisioning:
- Tenant requests compute capacity via portal or API.
- Orchestration validates available infrastructure using the digital twin.
- A 20,000 sq ft compute block (or fractional subset) is allocated.
- Optical connectivity is configured and activated.
- AI clusters are deployed and integrated with tenant security.
- SLA telemetry and monitoring begin automatically.
This model enables organizations to deploy hyperscale‑class AI infrastructure in minutes instead of months.