Datacenter Power Allocation for Dynamic Overhead Rebalancing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Datacenters face challenges in efficiently allocating power between computing systems, overhead systems, and backup power, leading to suboptimal utilization of input power and increased energy costs.

Innovation Solution

A power allocation system and method that utilize a power controller to dynamically reallocate overhead power to computing power based on computing load, thereby maximizing the utilization of input power for computing resources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If overhead power allocation is increased to ensure adequate power for overhead systems (lighting, HVAC, cooling), then reliability of overhead services is improved, but computing power allocation is reduced

Engineering Contradiction:

Improvereliability of overhead servicesVSAvoidcomputing power allocation

Core Design Contradiction:

ReliabilityVSPower

Solution Approach 1:

The power allocation system dynamically adjusts the split between overhead power allocation and computing power allocation based on real-time computing load conditions. When computing load increases, the system automatically increases computing power allocation by reducing overhead power allocation, and vice versa. This dynamic adjustment resolves the contradiction by making the power distribution flexible rather than fixed, allowing the system to adapt to varying operational requirements.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the power allocation parameters (the percentage split between overhead and computing power) based on computing load thresholds. When computing load exceeds a threshold, the system changes the allocation parameter to favor computing power; when load is below the threshold, it adjusts to favor overhead services. This parameter change approach allows the system to resolve the contradiction by adjusting operational parameters rather than physical infrastructure.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If computing power allocation is increased to maximize computing capacity, then productivity of computing resources is improved, but power for overhead systems is reduced

Engineering Contradiction:

Improvecomputing capacity utilizationVSAvoidpower for overhead systems

Core Design Contradiction:

ProductivityVSUse of energy by stationary object

Solution Approach 1:

The system dynamically adjusts overhead power allocation based on computing load. When computing productivity requirements are high (high computing load), the system automatically reduces overhead power allocation to maximize computing capacity. When computing load is low, it increases overhead power allocation to maintain overhead services. This dynamic behavior resolves the contradiction by making power distribution responsive to actual computing needs.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The power allocation system automatically monitors computing load and self-adjusts the power distribution without external intervention. When computing capacity utilization is high, the system self-adjusts to allocate more power to computing and less to overhead services, and vice versa. This self-service capability allows the system to resolve the productivity-overhead contradiction autonomously based on real-time conditions.

Inventive Principle:

Principle #25Self-service

3Ease of operation

If fixed power allocation is used between overhead and computing services, then ease of operation is improved, but adaptability to varying computing load is reduced

Engineering Contradiction:

Improvesimplicity of power allocationVSAvoidadaptability to computing load

Core Design Contradiction:

Ease of operationVSAdaptability or versatility

Solution Approach 1:

The system transitions from static fixed power allocation to dynamic adaptive power allocation. The power distribution automatically adjusts based on computing load thresholds, maintaining simplicity of operation through automated rule-based adjustment while gaining adaptability to varying computing demands. This resolves the contradiction by replacing manual reconfiguration needs with automated dynamic adjustment.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The power allocation system incorporates feedback from computing load monitoring to automatically adjust power distribution. When computing load changes, the system receives feedback and adjusts the power allocation accordingly, maintaining both ease of operation (through automation) and adaptability (through responsive adjustment). This feedback mechanism resolves the contradiction by creating a closed-loop system that adapts to conditions while remaining simple to operate.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS12314102B2Allocating power between overhead, backup, and computing power services

Publication Date: 2025.05.27 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

A power controller allocates input power to a datacenter between computing power for computing services, backup power, and overhead power for overhead systems. The power controller reallocates the overhead power and/or the backup power to the computing power. This may increase the overall utilization of the datacenter by allowing additional processing power of the servers to be used.