Multi-Processor Workload Redistribution for Power Management

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

High-density computing environments face significant power consumption and heat generation issues, making existing power-saving techniques, such as standby modes, ineffective for servers that must remain operational, especially in environments with limited air conditioning capacity.

Innovation Solution

A multi-processor system redistributes computing loads among processors, allowing underutilized processors to be throttled down or put into sleep mode, reducing power consumption by offloading workloads and adjusting processor speeds within configured tolerances.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If more processors are deployed to increase compute density, then computational capacity is improved, but power consumption and heat generation increase

Engineering Contradiction:

Improvecomputational capacityVSAvoidpower consumption

Core Design Contradiction:

ProductivityVSUse of energy by moving object

Solution Approach 1:

The system dynamically adjusts processor states based on real-time workload requirements. Processors transition between active and sleep modes according to actual computational needs, allowing the system to optimize the balance between computational capacity and power consumption rather than operating all processors at fixed capacity

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes operational parameters of processors by adjusting clock speeds and power states. By varying these parameters based on workload distribution, the system achieves high computational capacity when needed while minimizing power consumption during low-demand periods

Inventive Principle:

Principle #35Parameter changes

2Use of energy by moving object

If processors are shut down to save power, then power consumption is reduced, but system availability and reliability deteriorate

Engineering Contradiction:

Improvepower consumptionVSAvoidsystem availability

Core Design Contradiction:

Use of energy by moving objectVSReliability

Solution Approach 1:

The system segments processor operations into distinct active and sleep states, allowing individual processors to be powered down while maintaining system-level availability. This segmentation enables granular power management where not all processors need to be shut down simultaneously, preserving system reliability while reducing overall power consumption

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system implements feedback mechanisms that continuously monitor workload distribution and processor utilization. This feedback loop enables intelligent decision-making about which processors to power down and when, ensuring that system availability is maintained by keeping critical processors active while powering down idle ones

Inventive Principle:

Principle #23Feedback

3Use of energy by stationary object

If standby mode is implemented to save power, then individual device power savings are improved, but overall system power savings deteriorate

Engineering Contradiction:

Improveindividual device power savingsVSAvoidoverall system power savings

Core Design Contradiction:

Use of energy by stationary objectVSLoss of energy

Solution Approach 1:

The system merges the power-saving capabilities of individual processors into a coordinated system-wide strategy. By synchronizing processor state changes and workload distribution across multiple CPUs, the system achieves synergistic power savings that exceed the sum of individual processor savings, overcoming the limitation of standalone standby mode implementations

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS7739527B2System and method to enable processor management policy in a multi-processor environment

Publication Date: 2010.06.15 TAHOE RES LTD

AI summary

In some embodiments, the invention involves off-loading processor workloads to reduce power requirements of a multi-processor system. In one embodiment, a multi-processor system redistributes computing among the multiple processors and changes the state of one or more processors to sleep mode. In another embodiment, a multi-processor system throttles the processor speed of under-utilized processors to reduce power consumption. Other embodiments are described and claimed.