Worked example

Edge Compute 50–125 kW (Reference)

Compact 800V DC package for edge data / AI nodes.

Envelope: 50–125 kW (DC Edge)

Design-target edge node: dense DC distribution, storage for ride-through, metering, and open-architecture control — scaled to DC Edge capacity family.

Problem framing

Edge AI needs data-centre-class power quality and density at smaller scale and faster deployment than traditional electrical fit-out.

Architecture intent

Compact conversion to an 800V DC bus, modular distribution, and BESS for short-duration support.

Metering and EMS provide remote visibility — design-target stack, not a shipping SKU list.

What this is not

Not a hyperscale design. Not a UPS replacement warranty. Scale deliberately with the DC Edge family.

Energy flow

Stages in this reference

  1. 01Grid InteractionGrid-forming, grid-following, export limiting, Volt/VAR, frequency response, black start.
  2. 02ConversionAC ↔ DC conversion, voltage transformation — PCS, SST, rectifiers, inverters.
  3. 03DistributionDC power distribution — busway, cabling, termination, tap-off units.
  4. 04ProtectionDC arc protection, digital relays, circuit protection, isolation.
  5. 05LoadsEnergy consumption — EV charging, HVAC, motors, data centres, buildings, processes.
  6. 06ControlEnergy management, dispatch optimisation, forecasting, VPP participation.
  7. 07MeteringMeasurement at every point — grid edge, DC bus, load, charger. IEEE 2030.5-native.

Want a site-specific single-line conversation?

Bring constraints and capacity envelope — architecture first.