Data Center Power Management Using Predictive Analytics

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Data centers face challenges in balancing power efficiency with Service Level Agreement (SLA) compliance due to dynamic workloads and varying utilization rates, leading to inefficient energy consumption and cooling demands.

Innovation Solution

A system and method using predictive analytics to set power caps based on utilization rates and cooling capacity, adjusting CPU frequency and voltage to optimize power consumption while meeting SLA requirements, employing machine learning models for dynamic power management.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If servers are put in deep idle states to conserve energy, then power consumption is reduced, but restoration time increases and activation energy costs increase

Engineering Contradiction:
Improvepower consumptionVSAvoidrestoration time
Core Design Contradiction:
Use of energy by moving objectVSLoss of time

Solution Approach 1:

The patent implements dynamic power cap adjustment based on real-time utilization rates and predictive analytics. The system continuously adapts power caps between deep idle states and active states, transitioning servers dynamically rather than statically. This allows servers to remain in lower-power states longer while quickly recovering when needed, balancing energy savings with restoration speed requirements.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system uses predictive analytics to forecast future utilization rates and proactively adjusts power caps before actual demand changes occur. By anticipating future needs, the system can prepare servers in optimal states, avoiding the need for deep idle states when quick recovery is anticipated, thus reducing restoration time penalties.

Inventive Principle:
Principle #10Preliminary action

2Speed

If CPU frequency is increased to improve processing speed, then response time is reduced, but power consumption increases due to voltage increase

Engineering Contradiction:
Improveprocessing speedVSAvoidpower consumption
Core Design Contradiction:
SpeedVSUse of energy by moving object

Solution Approach 1:

The patent dynamically adjusts CPU frequency and voltage based on real-time utilization rates and predictive analytics forecasts. Rather than maintaining high fixed frequencies, the system continuously optimizes the frequency-voltage profile to match actual workload demands, reducing power consumption during low-utilization periods while maintaining performance during high-demand periods.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes multiple parameters including CPU frequency, voltage, and power caps simultaneously based on predictive analytics models. By coordinating these parameter changes together, the system achieves optimal performance-power tradeoffs that cannot be achieved by adjusting frequency alone, as the voltage-frequency relationship is dynamically optimized based on predicted workload patterns.

Inventive Principle:
Principle #35Parameter changes

3Use of energy by moving object

If power caps are reduced to save energy, then power consumption is reduced, but SLA compliance may be violated

Engineering Contradiction:
Improvepower consumptionVSAvoidSLA compliance
Core Design Contradiction:
Use of energy by moving objectVSReliability

Solution Approach 1:

The system continuously monitors actual utilization rates and SLA compliance metrics, using this feedback to adjust power caps dynamically. The predictive analytics model incorporates historical SLA data to forecast future compliance risks, allowing the system to adjust power caps in advance to prevent SLA violations while maximizing energy savings during periods when compliance risks are low.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system uses predictive analytics to forecast future utilization patterns and proactively adjusts power caps before SLA compliance issues arise. By anticipating future demand patterns and potential compliance risks, the system can set appropriate power caps in advance, ensuring SLA compliance is maintained while energy savings are maximized during stable periods.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4139773B1System and methods for server power management
Publication Date: 2026.01.21 ZUTA CORE LTD
  • EP4139773B1 patent drawingFigure 1
  • EP4139773B1 patent drawingFigure 2

AI summary

A system and methods are provided for improving power efficiency of a data center, including: acquiring training data including power caps, utilization rates, and a measure of Service Level Agreement (SLA) compliance of one or more computer servers of the data center; creating a model for determining power caps according to measured utilization rates of the one or more computer servers, wherein the determined power caps, when applied to the one or more computer servers, reduce power consumption and meet the measure of SLA compliance; and applying the model, according to subsequent data received during a second operating period, to determine a power cap to apply to the one or more computer servers, wherein the subsequent data includes a subsequent utilization rate of the one or more computer servers.