BlueEast RightScale™

Adaptive Infrastructure.
Not Assembled Components.

BlueEast RightScale™ is an adaptive infrastructure platform built for the rapidly evolving demands of distributed accelerated compute. Designed with both power-agnostic and compute-adaptive architecture, the platform scales in two critical dimensions — overall platform power capacity and rack-level compute density — without costly electrical or cooling infrastructure rebuilds.

10 MW
Per Node Capacity
48 VDC
Native Safe-Power Output
>97%
Wall-to-Rack Efficiency
<1.2
Target PUE
3–6 wks
Deployment Time

48VDC Safe-Power Distribution.
Power-Agnostic by Design.

Every other adaptive infrastructure platform still converts AC to DC at the rack, burning 5–8% of contracted power as heat before a single compute cycle begins. BlueEast RightScale™ eliminates this stage entirely, delivering stabilised −48 VDC directly via advanced busbar topology to every rack position.

The Big Wheel Poly-Draw platform accepts solar PV, compressed air energy storage (CAES), grid, and generator input simultaneously or independently — supporting virtually any primary energy strategy without infrastructure redesign. Power-agnostic architecture means operators are never locked into a single energy source.

With 6 PCUs and mesh cross-connects throughout, there is no single point of failure in the power distribution plane. Any node overload is automatically redistributed before it can escalate.

  • Ripple-free output — near-zero AC ripple (<0.1%) on the DC bus versus 1–5% in solid-state converters; stable voltage references, reduced EMI, and no ripple-induced thermal cycling of capacitor banks
  • Ringing eliminated — no switching nodes means self-oscillating ringing effects (parasitic LC resonance at MHz frequencies) cannot occur; not suppressed — physically absent
  • Surge & spike levelling — rotating mass passively absorbs grid-side surges, lightning transients, and load-side GPU power ramps through mechanical inertia before they reach the DC bus
  • Ground isolation — electro-mechanical isolation prevents AC input faults propagating to the DC distribution bus; positive-ground topology eliminates ground loops when blending sources with different ground references
  • Galvanic corrosion addressed — C110 ETP copper single-metal busbar family, silver-over-nickel barrier plating on contact surfaces, and G10 dielectric spacers interrupt corrosion at the material, plating, and isolation levels
  • Spontaneous overload protected — 6-PCU mesh auto-redistributes on any node overload; 48V DC operating voltage limits prospective fault current; per-circuit PDU protection provides the final barrier at rack level
  • EMP hardened — positive-ground topology and electro-mechanical isolation provide inherent protection; no active shielding required
  • Compute-adaptive scalability — platform power capacity and rack-level compute density both expand without requiring costly electrical or cooling redesigns
  • 500kW–1MW recovered per 10MW node versus AC-distribution competitors
  • Ready for 48V native next-gen miners and NVIDIA 800V GPU hardware today
Power Output
Output voltageStabilised −48 VDC
Node capacity10 MW
Wall-to-rack efficiency>97%
DC ripple<0.1% — near-zero
Ringing effectsAbsent — no switching nodes
Surge levellingPassive — mechanical inertia
Isolation & Protection
Ground topologyPositive-ground, isolated
Input isolationElectro-mechanical — AC faults cannot reach DC bus
Galvanic corrosionAddressed — Cu/Ag-Ni/G10 system
EMP protectionInherent — positive-ground topology
Redundancy & Overload
Architecture6-PCU mesh
Single points of failureZero
Overload responseAuto-redistributed — 6-PCU mesh
Fault current limitInherent — 48V DC topology
Final barrierPer-circuit monitored PDUs
Input Sources
Solar PV✓ Native
Grid✓ Native
Generator✓ Native
CAES✓ Native
Simultaneous blend✓ Supported

Three Cooling Modes.
One Adaptive Platform.

No other adaptive infrastructure platform supports immersion, direct-to-chip liquid cooling, and free-air/adiabatic from a single native architecture. BlueEast was designed from the ground up to handle any compute workload — AI/GPU clusters, mining ASICs, HPC nodes — without a platform change. This compute-adaptive thermal architecture supports staged density growth — from initial deployment through multi-megawatt expansion — without platform redesign.

Our direct-to-chip system captures 70–85% of IT heat at the source, targeting PUE below 1.2 in favorable climates. Immersion capability handles the highest-density ASIC configurations. Rear-door heat exchangers provide a fallback layer for mixed environments.

  • Engineered for racks exceeding 100 kW — highest per-rack density in the competitive set
  • 70–85% heat capture at chip level via direct-to-chip liquid cooling
  • Immersion-ready for ASIC workloads — Shell fluid compatible
  • Free-air and adiabatic modes for edge and low-density deployments
  • Rear-door heat exchanger fallback for hybrid environments
Cooling Performance
Target PUE<1.2 (favorable climates)
Liquid heat capture70–85% of IT load
Max per-rack density100+ kW
Supported Cooling Modes
Direct-to-chip (DLC)✓ Native
Single-phase immersion✓ Native
Free-air✓ Native
Adiabatic✓ Native
Rear-door HX✓ Fallback

Deployment-Optimized Control.
Sub-Second Response.

BlueEast modules ship with ESCOlutions energy management fully integrated — not bolted on. The system maintains a real-time digital twin of every module, enabling predictive maintenance before failures occur and automated failover orchestration without operator intervention.

Sub-second telemetry loop closure means the control system responds to load transients faster than any traditional SCADA implementation. Workload-aware dispatch automatically re-routes power delivery based on active compute demand.

  • ESCOlutions platform — AI-driven predictive maintenance and anomaly detection
  • Sub-second control loop closure for transient load response
  • Real-time digital twin — full module state visible at all times
  • Open API for DCIM/ITSM integration — no vendor lock-in
  • Automated failover orchestration — no operator intervention required
  • Workload-aware power dispatch — optimizes distribution to active compute
Control System
PlatformESCOlutions
Telemetry loopSub-second closure
Digital twinReal-time
APIOpen — DCIM/ITSM
AI Capabilities
Predictive maintenance
Anomaly detection
Workload-aware dispatch
Automated failover
Technical Consultation

Ready to go deeper?

BlueEast RightScale™ is engineered for organizations that require infrastructure scaling intelligently over time — without costly redesign cycles or stranded infrastructure investment. Request a full technical brief or arrange a walkthrough with our engineering team.