Workload Assignment Using Power Utilization Index

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Data centers face significant power consumption challenges due to growing server volumes, with existing technologies lacking reliable estimates of energy consumption per server for given workloads, making it difficult to optimize power usage across diverse server configurations and locations.

Innovation Solution

The development of a power utilization index, or Portable Energy Consumption (PEC) metric, which estimates energy consumption per unit of workload, allowing for predictive modeling of future power usage and strategic workload assignment across servers, racks, aisles, and data centers to minimize overall power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If server volumes grow to meet increasing data processing demands, then data center capacity and processing power improve, but overall power consumption increases significantly

Engineering Contradiction:

Improvedata processing capacityVSAvoidpower consumption

Core Design Contradiction:

ProductivityVSUse of energy by moving object

Solution Approach 1:

The patent changes the parameter of workload assignment from random or static allocation to dynamic assignment based on measured power consumption metrics. By continuously monitoring and adjusting workload distribution parameters according to real-time power consumption data, the system optimizes the balance between processing capacity and energy usage.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent implements a feedback mechanism where power consumption is measured, compared against predicted values, and used to adjust future workload assignments. The system calculates differences between predicted and actual power consumption, then uses this feedback to make informed decisions about where to assign new workloads, creating a closed-loop control system that continuously improves energy efficiency.

Inventive Principle:

Principle #23Feedback

2Adaptability or versatility

If diverse server configurations from different manufacturers and generations are deployed to meet varying workload requirements, then system versatility and adaptability improve, but measurement and comparison of power consumption becomes difficult

Engineering Contradiction:

Improveserver configuration diversityVSAvoidpower consumption comparison accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent creates a universal power consumption metric that works across diverse server configurations from different manufacturers and generations. By developing a standardized measurement approach that can be applied uniformly to any server, the system enables accurate comparison and optimization across heterogeneous infrastructure without requiring configuration-specific models.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent transforms power consumption measurements into a standardized parameter format that allows direct comparison across different server types. By normalizing the measurement data and creating a common metric framework, the system eliminates the complexity of comparing diverse configurations and enables straightforward optimization decisions.

Inventive Principle:

Principle #35Parameter changes

3Ease of operation

If traditional workload assignment methods are used without considering non-data processing overhead, then assignment simplicity is maintained, but overall power consumption optimization is lost

Engineering Contradiction:

Improveworkload assignment simplicityVSAvoidnon-data processing overhead power consumption

Core Design Contradiction:

Ease of operationVSUse of energy by moving object

Solution Approach 1:

The patent extracts and isolates the non-data processing overhead component from total power consumption by comparing predicted power consumption (based on workload) with actual measured power consumption. This extracted difference represents the overhead power consumption, which is then used as a separate metric for optimization decisions independent of the data processing workload itself.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces an intermediary metric (the difference between predicted and actual power consumption) that mediates between simple workload assignment and comprehensive power optimization. This intermediary measurement allows the system to account for overhead consumption without requiring complete redesign of the workload assignment process.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS10310912B2Assigning data processing workloads in data centers based on non-data processing overhead

Publication Date: 2019.06.04 CA TECH INC

AI summary

Data processing workloads are selectively assigned within a data center and/or among data centers based on non-data processing overhead within the data center and/or among the data centers. Power consumption of a rack including servers is predicted based on data processing demands that are placed on the servers for a given data processing workload, and power consumed by the rack is measured when the servers are performing the given data processing workload. A metric of power consumed by the rack for non-data processing overhead is derived based on a difference between results of the predicting and the measuring. A future data processing workload is selectively assigned to the rack based on the metric of power of power consumed by the rack for the non-data processing overhead. Assignment may also take place at an aisle and/or data center level based on these metrics.