Network Power Budget Optimized Application Scheduling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Multi-node computer system networks face inefficiencies in power consumption and resource allocation due to the lack of effective methods for optimizing application processes and interactions across different nodes and pathways, leading to increased power consumption and resource wastage.

Innovation Solution

A network optimization system that uses a processor to identify clusters, network devices, applications, and their interactions, determining hardware and application constraints to generate modifications that optimize network performance and power consumption by reassigned applications and pathways, leveraging machine learning models for predictive rescheduling and resource allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If applications are scheduled across multiple nodes and pathways in a multi-node network, then application interaction and distribution are improved, but power consumption increases

Engineering Contradiction:

Improveapplication interaction efficiencyVSAvoidnetwork power consumption

Core Design Contradiction:

ProductivityVSUse of energy by moving object

Solution Approach 1:

The system dynamically changes scheduling parameters such as application-node assignments and pathway selections based on real-time power consumption metrics and performance requirements. The optimization engine adjusts these parameters to find the optimal balance between application interaction efficiency and power consumption, transitioning from static to dynamic scheduling configurations.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent implements dynamic scheduling where application assignments and network pathways are not fixed but can be reassigned in real-time. The system continuously monitors performance and power consumption, then dynamically adjusts the scheduling configuration to adapt to changing conditions, allowing the network to optimize its operation state based on current demands.

Inventive Principle:

Principle #15Dynamics

2Adaptability or versatility

If more pathways are used for application interactions, then network flexibility is improved, but resource allocation efficiency deteriorates

Engineering Contradiction:

Improvenetwork flexibilityVSAvoidresource allocation efficiency

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The system does not attempt to optimize all possible pathways simultaneously, but rather selects and optimizes a subset of pathways that provide sufficient flexibility for application interactions. The optimization engine identifies and focuses on the most critical application interactions and their corresponding pathways, avoiding the complexity of managing all possible routing options.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The patent applies different optimization strategies to different parts of the network based on local requirements. Rather than applying a uniform optimization approach across the entire network, the system tailors scheduling decisions to specific application interactions and their unique requirements, allowing each region of the network to be optimized according to its specific needs.

Inventive Principle:

Principle #3Local quality

3Speed

If application scheduling is optimized for performance, then application process speed is improved, but power consumption increases

Engineering Contradiction:

Improveapplication process speedVSAvoidpower consumption

Core Design Contradiction:

SpeedVSUse of energy by moving object

Solution Approach 1:

The system implements periodic optimization cycles where scheduling decisions are reviewed and adjusted at regular intervals. Rather than continuously optimizing at maximum intensity, the system performs optimization iterations periodically, allowing performance-critical applications to maintain high speed while reducing overall power consumption during non-critical periods.

Inventive Principle:

Principle #19Periodic action

Solution Approach 2:

The optimization engine proactively identifies and prevents suboptimal scheduling decisions before they occur. By predicting future power consumption patterns and performance requirements, the system makes preemptive scheduling adjustments that avoid the need for high-power operations, thereby reducing overall power consumption while maintaining acceptable performance levels.

Inventive Principle:

Principle #9Preliminary anti-action

Data Source

PatentUS12166639B2Network power budget optimized scheduling of applications

Publication Date: 2024.12.10 RED HAT INC

AI summary

The present disclosure provides new and innovative systems and methods for optimizing networks for the scheduling and execution of applications. An example method includes identifying a plurality of clusters network devices associated with a network. For each cluster and for each network device, a respective set of hardware constraints are determined. Furthermore, applications associated with each cluster may be determined. For each application associated with each cluster, application interactions and a set of application constraints may be determined. The application interactions may occur between applications associated with different clusters and may include a first application interaction between a first application associated with a first cluster and a second application associated with a second cluster. The method further includes generating a modification to the first application interaction, wherein the modification optimizes the network based on an optimization criteria.