Chassis Power Budgeting for Modular Servers

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing information handling systems lack effective power budgeting, leading to immediate power draw by server nodes without checks, potential chassis shutdowns due to exceeded power supply capacity, and inability to calculate power limits, resulting in poor performance and safety issues.

Innovation Solution

A method and system that determine redundancy configuration, power capacity, and number of modular information handling systems to set chassis-level maximum transient and average power limits, communicating these limits to baseboard management controllers for sled-level power requirement checks, thereby preventing overconsumption and ensuring system performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If server nodes draw power immediately on power button press without power budget check, then power availability and responsiveness are improved, but system reliability deteriorates due to potential chassis shutdown when power consumption exceeds supply capacity

Engineering Contradiction:

Improvepower response timeVSAvoidsystem stability

Core Design Contradiction:

SpeedVSReliability

Solution Approach 1:

The system performs preliminary power budget checks before allowing server nodes to draw power. The chassis manager firmware calculates available power capacity based on redundancy configuration and communicates power limits to baseboard management controllers, which then verify power requirements before power button press, preventing oversubscription while maintaining fast response through pre-calculated power headroom.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements continuous feedback loops where baseboard management controllers monitor actual power consumption against allocated budgets and report back to the chassis manager. When power limits are approached, the system dynamically adjusts power allocation and notifies relevant components, enabling real-time power management that maintains both responsiveness and stability.

Inventive Principle:

Principle #23Feedback

2Extent of automation

If existing chassis manager firmware supports power budgeting, then power management capability is improved, but customer experience deteriorates due to throttling and poor performance

Engineering Contradiction:

Improvepower management automationVSAvoidsystem performance

Core Design Contradiction:

Extent of automationVSProductivity

Solution Approach 1:

The system dynamically adjusts power allocation based on actual system conditions, workload requirements, and available power headroom. Instead of static power limits that cause throttling, the chassis manager continuously recalculates power budgets and communicates updated limits to baseboard management controllers, allowing the system to adapt to changing conditions and maintain optimal performance within power constraints.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes power management parameters dynamically, including transient power limits, average power limits, and power capping boundaries. By adjusting these parameters based on redundancy configuration, power capacity, and number of installed systems, the firmware optimizes the balance between power safety and performance, eliminating the throttling issues of static approaches.

Inventive Principle:

Principle #35Parameter changes

3Device complexity

If existing approaches lack power budgeting, then system simplicity is improved, but ability to calculate power limits deteriorates, preventing fast throttle boundary and node lower boundary calculation

Engineering Contradiction:

Improvepower management complexityVSAvoidpower limit calculation precision

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The system segments power management into distinct functional layers: chassis-level power capacity calculation, sled-level power budget allocation, and node-level power limit enforcement. Each layer handles specific calculations (chassis manager computes total capacity and transient limits, baseboard management controllers compute node lower boundaries and enforce power caps), distributing complexity across modular components while enabling precise power limit calculations that would be impossible in a monolithic simple system.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11249533B2Systems and methods for enabling power budgeting in an information handling system comprising a plurality of modular information handling systems

Publication Date: 2022.02.15 DELL PROD LP

AI summary

A method may include, in a system comprising a plurality of power supply units, determining a redundancy configuration of the plurality of power supply units, determining a power capacity of the plurality of power supply units, determining a number of modular information handling systems installed in the system, determining a chassis-level maximum transient power limit for each of the modular information handling systems based on the redundancy configuration, the power capacity, and the number of modular information handling systems, determining a chassis-level maximum average power limit for each of the modular information handling systems, and communicating the chassis-level maximum transient power limit and chassis-level maximum average power limit to a respective baseboard management controller of each of the modular information handling systems, wherein each respective baseboard management controller is configured to determine if sled-level power requirements of its respective information handling system are within the chassis-level maximum transient power limit and chassis-level maximum average power limit.