Server Load Adjustment Using Power Ratios for Energy Efficiency

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In enterprise environments, determining performance and power measurements for individual workloads is challenging due to the separation of execution spaces, making it difficult to optimize energy efficiency and carbon footprint.

Innovation Solution

A system that measures power consumption of servers and their sub-systems, identifies benchmark workloads associated with the current workload, and adjusts server load to achieve optimal energy efficiency by migrating jobs or applications based on power characteristics without requiring software integration.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If standard tools are used to measure performance and power in HPC environments, then measurements can be obtained for single-user-node allocation and bulk workloads, but it becomes challenging to determine measurements for every single workload in enterprise environments with shared resources and separated execution spaces

Engineering Contradiction:

Improveworkload performance and power measurementsVSAvoidenterprise environment execution space separation

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces a measurement intermediary system that sits between the separated execution spaces and the analysis tools. This intermediary captures power consumption data at the hardware level (using sensors on power supply units) and performance data (using instrumentation libraries), then correlates them through a common identifier (VM or container ID). This mediator approach resolves the measurement challenge by providing a unified view despite execution space separation.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces traditional mechanical/software-based measurement approaches (which require direct access to execution spaces) with an electrical/power-based measurement system. By measuring power consumption at the electrical level (using power sensors and metering devices) rather than trying to measure through software interfaces in separated execution spaces, the system achieves accurate workload-level measurements without being hindered by execution space separation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Adaptability or versatility

If virtual machines are separated into individual execution spaces, then each VM can run unrelated workloads independently, but it becomes difficult to determine power and performance measurements for every single workload

Engineering Contradiction:

ImproveVM workload independenceVSAvoidworkload-level power and performance data

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent segments the measurement system into distinct components: power measurement devices at the hardware level, performance measurement tools at the software level, and a correlation layer that links them through workload identifiers. This segmentation allows independent measurement of each VM's power consumption and performance metrics while maintaining the ability to correlate them, thus preserving both VM independence and measurement precision.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent implements feedback mechanisms where measurement data from separated execution spaces is continuously collected, correlated, and fed back to the management system. This feedback loop enables real-time monitoring and analysis of workload-level power and performance metrics even in separated execution environments, allowing for dynamic optimization and reporting.

Inventive Principle:

Principle #23Feedback

3Productivity

If data centers grow in size and number to meet computing demands, then more resources become available, but energy efficiency and carbon footprint become greater challenges

Engineering Contradiction:

Improvecompute resource capacityVSAvoiddata center energy consumption

Core Design Contradiction:

ProductivityVSLoss of energy

Solution Approach 1:

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent enables dynamic adjustment of workload parameters (such as load balancing decisions, resource allocation, and scheduling) based on real-time power and performance measurements. By changing operational parameters according to measured efficiency metrics, the system optimizes energy consumption while maintaining productivity, addressing the contradiction between growing compute capacity and managing energy loss.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20260029832A1Optimizing energy efficiency of server load based on power measurements and characteristics

Publication Date: 2026.01.29 HEWLETT PACKARD ENTERPRISE DEV LP

AI summary

A system determines a plurality of benchmark workloads including a compute-heavy workload, a memory-heavy workload, and an input/output (I/O)-heavy workload. The system obtains, for a respective benchmark workload, a benchmark power measurement associated with a benchmark system. The benchmark power measurement indicates a target efficiency threshold for the respective benchmark workload. The system measures power characteristics for a current workload on a computing device. The power characteristics comprise current power measurements associated with the computing device, processing components of the computing device, memory components of the computing device, and I/O components of the computing device. The system identifies, based on a ratio between two of the power characteristics, a benchmark workload most closely associated with the current workload. The system optimizes operation of the computing device by adjusting the current workload until an overall power consumption of the computing device reaches the target efficiency threshold for the identified benchmark workload.