Cluster Manager for Dynamic Process Migration and Power Optimization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing embedded systems face challenges in optimizing power efficiency and load balancing across multiple processors in clusters, particularly in dynamic environments where applications require efficient distribution and processing, while existing solutions often rely on complex hardware or neglect power consumption issues.

Innovation Solution

A software-based multi-processing system with a cluster manager that dynamically allocates processes and adjusts power settings across CPUs, using feedback from software-monitored CPU loads to optimize power usage and balance load, without requiring dedicated hardware for cache coherency or complex reconfiguration.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If hardware arrangements are used to enable cache coherency and inter-processor communication, then communication efficiency between processors is improved, but device complexity and manufacturing cost increase

Engineering Contradiction:

Improveinter-processor communication efficiencyVSAvoidhardware complexity

Core Design Contradiction:

SpeedVSDevice complexity

Solution Approach 1:

The patent replaces hardware-based cache coherency mechanisms with a software-based solution. The cluster manager software coordinates process migration and communication between processors, eliminating the need for dedicated hardware support for cache coherency while maintaining communication efficiency through software-managed process movement and shared memory access.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces a cluster manager software agent as an intermediary between processors. This software component coordinates communication, manages process migration, and balances load across processors, replacing direct hardware-based inter-processor communication mechanisms with a software-mediated approach that reduces hardware complexity.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If dedicated hardware support is implemented for cache coherency, then cache synchronization efficiency is improved, but manufacturing cost and device complexity increase

Engineering Contradiction:

Improvecache synchronization efficiencyVSAvoidmanufacturing cost

Core Design Contradiction:

ProductivityVSEase of manufacture

Solution Approach 1:

The patent substitutes hardware-based cache coherency support with a software-based approach. The cluster manager software handles cache synchronization by managing process migration and coordination, eliminating the need for expensive dedicated hardware circuits while maintaining synchronization efficiency through software-controlled mechanisms.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Adaptability or versatility

If software-based process migration is used for load balancing, then adaptability and ease of configuration are improved, but processing overhead and complexity increase

Engineering Contradiction:

Improveload balancing flexibilityVSAvoidsoftware complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The cluster manager software performs multiple functions including process migration, load balancing, communication coordination, and power management. By consolidating these diverse functions into a single software agent, the system achieves high adaptability without proportionally increasing complexity, as the same software infrastructure supports multiple operational requirements.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The cluster manager software automatically monitors system state and performs process migration and load balancing decisions without external intervention. This self-service approach allows the system to adapt dynamically to changing conditions while keeping operational complexity manageable through automated decision-making algorithms.

Inventive Principle:

Principle #25Self-service

4Productivity

If multiple processors are activated to handle complex applications, then computational power and productivity are improved, but power consumption increases

Engineering Contradiction:

Improvecomputational powerVSAvoidpower consumption

Core Design Contradiction:

ProductivityVSUse of energy by moving object

Solution Approach 1:

The patent implements dynamic processor activation and deactivation based on real-time load conditions. The cluster manager monitors CPU utilization and selectively activates or deactivates processors to match actual computational demands, ensuring high productivity when needed while minimizing power consumption during low-utilization periods through dynamic power state transitions.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system employs feedback mechanisms where the cluster manager continuously monitors processor load and power consumption, then adjusts processor activation states accordingly. This closed-loop control enables the system to optimize the trade-off between computational power and energy usage by responding to actual system conditions rather than operating in fixed states.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS8321693B2Parallel processing method and system, for instance for supporting embedded cluster platforms, computer program product therefor

Publication Date: 2012.11.27 STMICROELECTRONICS SRL

AI summary

A multi-processing system-on-chip including a cluster of processors having respective CPUs is operated by: defining a master CPU within the respective CPUs to coordinate operation of said multi-processing system, running on the CPU a cluster manager agent. The cluster manager agent is adapted to dynamically migrate software processes between the CPUs of said plurality and change power settings therein.