Tag Archive for: 800 VDC

data centerApplication Notes, Data Center

Manageable Power Converters: PMBus, Redfish and SCMI in the 800 VDC AI Data Centre

Application Note · Data Center

Manageable Power Converters: PMBus, Redfish and SCMI in the 800 VDC AI Data Centre

Which protocols sit between the data centre manager and the power converter, where each one fits, and what the converter firmware needs to provide.

For decades, the relationship between a power converter and the system around it was almost one-way: the load demanded power and the converter supplied it. Signalling was limited to an enable signal, a fault signal and, at most, a voltage setpoint.

In the AI rack, that model is no longer enough. As rack power grows towards 800 VDC distribution architectures (covered in The 800 VDC revolution), the data centre manager needs to know at all times how much power each conversion stage is delivering, how much thermal margin it has and whether there are any alarms. The converter stops being a passive element and becomes a managed node.

This note explains which protocols are involved in that management chain, where each one fits and which firmware capabilities a converter needs to integrate into it.

The management chain: from the data centre manager to the converter

No single protocol links the data centre manager to the converter. Management is organised in layers, and each layer uses a different standard:

Layer Typical protocol Function
Data centre manager ↔ rack controller / BMC Redfish (DMTF), REST API over HTTPS Inventory, telemetry, alarms and power policies at facility scale
Rack controller ↔ power shelf / BBU shelf Modbus over RS-485 or CAN (common in Open Rack V3) Reading measurements and status from power and backup shelves
Controller ↔ power supply or converter PMBus Telemetry and control of each conversion unit
Inside the compute node Arm SCMI Power and performance management of the silicon, including power capping

For the converter manufacturer, the practical consequence is clear: the equipment must speak PMBus comprehensively and expose the information that the upper layers will translate into Redfish, either directly or through the shelf or rack controller.

PMBus: the converter’s own language

PMBus is the open digital communication standard for power supplies and converters. Its current published revision is 1.5. It defines a common set of commands to read measurements and status and to configure the equipment, for example:

  • Measurements: READ_VIN, READ_IOUT, READ_POUT, READ_TEMPERATURE_1.
  • Status and alarms: STATUS_WORD and the specific status registers (voltage, current, temperature).
  • Configuration: warning and fault thresholds, and operating setpoints.

A converter with a complete, well-documented PMBus implementation simplifies integration: the rack controller can read its data without custom development.

Redfish: the facility view

Redfish is the DMTF standard for infrastructure management through a REST API with JSON data. For power it defines specific schemas, including PowerSubsystem, PowerSupply and PowerSupplyMetrics, as well as a generic sensor model.

The converter does not always need to implement Redfish itself. Typically, the shelf or rack controller collects the PMBus or Modbus data from each unit and publishes it in Redfish. What is essential is that the converter provides all the information those schemas expect: power, voltages, currents, temperatures, efficiency and status.

SCMI: where it really fits

Arm SCMI (System Control and Management Interface) is a standard, transport-independent interface between system software and the platform control firmware. Its natural scope is inside the compute node, not the link to the power supply. It defines protocols for discovery (Base), power domains, system power management, performance, clocks, sensors, reset, voltage and power capping (Powercap).

The Powercap protocol includes, among others, the POWERCAP_CAP_GET and POWERCAP_CAP_SET commands to read and set the power limit of a domain, and POWERCAP_MEASUREMENTS_GET to read average power. Version 4.0 of the specification, still in draft, adds Concurrent Power Limits (CPL) per domain and a telemetry protocol.

How does this relate to the converter? Power capping is executed by the silicon through SCMI, but the decision to cap depends on telemetry from the whole power chain, including the converter: if its thermal or power margin runs out, the manager can reduce compute consumption before the conversion stage operates outside its safe operating area.

Closed-loop thermal control

In liquid-cooled AI racks, thermal control is usually organised in two nested loops. The inner loop, in the coolant distribution unit (CDU), keeps the coolant temperature close to its setpoint by acting on pumps and valves. The outer loop, in the manager, adjusts that setpoint and the power limits according to the thermal telemetry from the racks.

To take part in the outer loop, the converter must publish the temperature of its power semiconductors, the temperature of its magnetic components and its instantaneous efficiency. With that data, the manager can decide whether there is margin for a transient increase in demand or whether power should be limited to protect the equipment’s reliability.

What the converter firmware must provide

  • Complete PMBus telemetry: input and output voltage and current, power, internal temperatures and status registers.
  • Interface to the rack controller in the protocol specified by the integrator, for example Modbus over RS-485 or CAN.
  • Information compatible with the Redfish power model, published directly or through the controller.
  • Configurable limits and derating, with event and alarm notification.
  • Telemetry latency consistent with the control dynamics defined by the integrator.
  • Controlled, traceable firmware updates.
Operators building on open architectures are likely to make this management capability a purchasing criterion for power conversion equipment, just as EN 50155 conformity is today for on-board railway equipment.

References

  • Arm — System Control and Management Interface (SCMI), specification DEN0056.
  • DMTF — Redfish: PowerSubsystem, PowerSupply and PowerSupplyMetrics schemas.
  • PMBus — Power Management Bus Specification, revision 1.5.
  • Open Compute Project — Open Rack V3 specifications.

About Premium PSU

Premium PSU (Premium SA) designs and manufactures custom power supplies and power converters in Barcelona, with over 40 years of experience in railway, energy, industrial, data centre and defence applications.

Talk to our engineering team

Discuss your power conversion requirements with Premium PSU.

Contact us

The DC Power Standards Map for European AI Data CentresApplication Notes, Data Center

The DC Power Standards Map for European AI Data Centres

Standards · Data Center

The DC Power Standards Map for European AI Data Centres

Five principles, three architectures and an international alliance: who is writing the LVDC rules for the 800 VDC AI data centre, and what they mean for power electronics manufacturers.

The transition to low-voltage direct current as the distribution architecture of the European AI data centre is not only a hardware change. It is a standards and ecosystem change being built in real time, with the active participation of the sector’s principal industrial actors. The OCP EMEA Summit 2026, held in Barcelona in April 2026, was where that process took its most visible step: the approval of the OCP white paper on Low Voltage DC Power Distribution as a contribution to the Open Data Center Ecosystem for AI, and the presentation of Schneider Electric’s five principles for 800 VDC adoption.

This article maps the emerging standards territory — who is writing the rules, what the rules are, and what the implications are for European power electronics manufacturers that want to take part in this market.

OCP DCF Power Distribution LVDC: the canonical document

The OCP Foundation’s DCF Power Distribution LVDC white paper (version 1.0) was approved on 29 April 2026, the first day of the OCP EMEA Summit in Barcelona; OCP has since published version 1.1.0. Its co-leadership corresponds to Eaton and ABB, the volunteer co-leads of the OCP Power Distribution Project.

The document’s operational definition delimits the perimeter of the entire conversation: LVDC (Low Voltage Direct Current) is defined as direct voltage equal to or below 1,500 VDC. The 800 V bus — the ±400 V of the Mt. Diablo specification — falls comfortably within that range. The document specifically addresses LVDC distribution as the interface between incoming Medium Voltage AC (MVAC, currently in the facility grey space) and rack compute in the IT white space.

Of particular relevance to power electronics manufacturers is its treatment of MVAC-to-LVDC conversion — either via a Solid State Transformer (SST) or via a combination of a conventional MV/LV transformer with an isolated or non-isolated AC/DC converter. These are exactly the stages of the sidecar in Architecture 1 (LV input) and of the sidecars in Architectures 2 and 3 (MV input) of the Schneider Electric roadmap.

Schneider Electric WP213: five principles for adoption

Schneider Electric’s white paper 5 Principles for 800 VDC in AI Data Centers: Rack-level Architectures as the Immediate Enabler — published in March 2026, is the reference document of one of the world’s largest energy management manufacturers on the transition to 800 VDC. It identifies 800 VDC power racks — explicitly called sidecars in the abstract — as the immediate enabler of the transition.

The five principles the document proposes are the axes on which designers and operators should evaluate and adopt 800 VDC architectures:

  • Architectural variations. There is no single 800 VDC architecture. The document distinguishes between localised distribution (sidecar), centralised distribution with LV converters and centralised distribution with SST — each with different time horizons, scales and levels of complexity.
  • Ecosystem availability. The maturity of components — SiC rectifiers, WBG DC/DC shelves, connectors, manifolds, protection gear — is the variable that determines when and at what scale each architecture can be deployed.
  • Protection coordination and earthing. The transition to 800 VDC introduces specific challenges in arc fault protection and earthing schemes — 2-wire vs. 3-wire DC — that call for standards review and dedicated protection coordination studies for high-voltage DC.
  • Energy storage integration. Backup systems at 800 V change the sizing logic of the BBU and open the door to LIC or supercapacitor solutions connected directly to the DC bus, without an intermediate conversion stage.
  • Operational execution. Operator training, service procedures in 800 VDC installations, fault management and diagnostic tools are a matter of deep operational change that is not solved by the choice of hardware alone.

The OCP × Current/OS alliance: an open standard for DC

The third vector of the standards ecosystem is the alliance between the OCP Foundation and Current/OS, formalised in December 2025. Current/OS is an independent, non-profit global partnership that promotes open standards for interoperable DC microgrids and is developing a DC standard and DC-certified products for buildings and installations.

The alliance was promoted by the volunteer co-leads of the OCP Power Distribution Project — precisely Eaton and ABB, the same companies co-leading the LVDC white paper. Its stated objective is to combine OCP’s experience in open data centre hardware with the Current/OS standards for interoperable DC, creating what the announcements described as a vendor-neutral operating system for DC-powered data centres.

In addition, on 29 April 2026 — the opening day of the Summit — the OCP Foundation announced a new alliance with the Electric Power Research Institute (EPRI), aimed at integrating the dynamic management of AI data centres with grid services such as demand response, virtual power plants and frequency regulation.

IEC and the European standards landscape

The IEC landscape for LVDC is under active construction, with three standards families of direct relevance to the sidecar. IEC 62477 (Power Electronic Converter Systems) establishes the general safety requirements for power electronic converters, including the DC/DC conversion stages of the sidecar. IEC 62109 (Safety of Power Converters for Use in Photovoltaic Power Systems) provides a precedent for DC converter safety testing methodologies that can be applied by analogy to the data centre context. IEC 60364-8-1 (functional aspects — energy efficiency of low-voltage electrical installations) sets the energy-efficiency framework for the low-voltage installations that host this infrastructure.

Arc fault protection in DC systems above 400 V is one of the least mature areas of the standards landscape for data centre applications. This is both a regulatory risk — installations may require case-by-case assessment by notified bodies — and an opportunity for manufacturers that develop and qualify arc fault protection solutions for the 800 VDC sidecar.

Implications for European power electronics manufacturers

The standards landscape described above has five direct implications for European power electronics manufacturers targeting the data centre market:

  • Mt. Diablo compliance is likely to become a commercial requirement, so component qualification should be planned early.
  • Arc fault protection is an open technical and regulatory problem in which early movers can make a difference.
  • Converter-level telemetry will be required at converter level by operators building on the OCP LVDC architecture.
  • High availability requirements for inference data centres impose MTBF and redundancy requirements consistent with railway and defence-grade power electronics.
  • European traceability requirements under the EU Chips Act and the Cyber Resilience Act will progressively require documented supply chain provenance for critical infrastructure components, including data centre power electronics.

References

  • Open Compute Project Foundation — DCF Power Distribution LVDC white paper (v1.0 approved 29 April 2026; v1.1.0 published subsequently).
  • Schneider Electric — 5 Principles for 800 VDC in AI Data Centers (March 2026).
  • Open Compute Project Foundation and Current/OS — strategic alliance announcement (December 2025).
  • Open Compute Project Foundation and EPRI — alliance announcement (29 April 2026).
  • IEC 62477-1:2022 — Safety requirements for power electronic converter systems and equipment.
  • IEC 60364-8-1:2019 — Low-voltage electrical installations, Part 8-1: Functional aspects, Energy efficiency.
About Premium SA — Premium SA is a Barcelona-based designer and manufacturer of custom power converters for railway, energy, industrial, data centre and defence applications. Founded in 1981, the company brings over 40 years of experience in high-reliability power electronics.

Power designed for your application

Talk to our engineering team about power conversion for 800 VDC data centre architectures.

Contact the technical team