Server Workload Allocation for Data Center Cooling Efficiency

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Data centers face significant economic losses due to server shutdowns caused by heat overload, and existing cooling solutions, such as CRACs, consume substantial power and are not optimally configured to handle the concentrated heat generated by densely packed server computers.

Innovation Solution

A method and arrangement that allocate processing tasks to servers based on the location and efficiency characteristics of air conditioning units, ensuring that heat generated by processing tasks is distributed efficiently among air conditioning units, optimizing their operational efficiency levels.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If CRACs are used to provide enhanced cooling power in data centers, then cooling reliability is improved, but power consumption increases significantly

Engineering Contradiction:

Improvecooling reliabilityVSAvoidpower consumption

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system dynamically adjusts CRAC operation by allocating processing tasks based on real-time thermal conditions and CRAC efficiency characteristics. The load allocation is not static but adapts to changing thermal environments, allowing the system to maintain reliable cooling while optimizing power consumption through dynamic task migration between servers based on which CRAC units are most efficient at any given moment

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes operational parameters by considering CRAC efficiency characteristics and thermal proximity when allocating processing tasks. Instead of uniform task distribution, the system modifies task allocation parameters based on thermal conditions and CRAC performance metrics, thereby achieving more efficient cooling with reduced power consumption

Inventive Principle:

Principle #35Parameter changes

2Productivity

If servers are densely packed to increase processing capacity, then productivity is improved, but heat generation increases causing server shutdowns

Engineering Contradiction:

Improveprocessing capacityVSAvoidheat overload

Core Design Contradiction:

ProductivityVSObject-affected harmful factors

Solution Approach 1:

The system applies local quality by considering the thermal environment of specific server locations when allocating tasks. Servers are not treated uniformly but are assigned tasks based on their local thermal conditions and proximity to efficient CRAC units. This localized approach ensures that servers in cooler zones can handle more intensive workloads while those in warmer zones receive fewer or lighter tasks, preventing heat overload while maintaining overall productivity

Inventive Principle:

Principle #3Local quality

3Power

If CRACs are operated at high capacity to handle concentrated heat, then cooling performance is improved, but power consumption increases

Engineering Contradiction:

Improvecooling powerVSAvoidpower consumption

Core Design Contradiction:

PowerVSUse of energy by moving object

Solution Approach 1:

The system applies partial action by allocating processing tasks selectively based on CRAC efficiency characteristics rather than maximizing CRAC capacity utilization. Instead of operating all CRACs at high capacity, the system uses only the portion of cooling capacity that is most efficient at any given time, tasking servers based on which CRACs are operating at optimal efficiency points, thereby reducing overall power consumption while maintaining adequate cooling performance

Inventive Principle:

Principle #16Partial or excessive action

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach reduces power consumption by efficiently distributing heat loads among air conditioning units, thereby minimizing the risk of server shutdowns and optimizing cooling performance in data centers.

Implementation Method 1

The dense packing of the server computers results in the generation of a large amount of heat in a localized area. The data center must be cooled in a reliable manner

Methodology Applied

Scientific EffectThermal conduction: Conduction (thermal)

Implementation Method 2

specialized cooling units have been developed for implementation directly in data centers... CRACs have been employed as a result of the fact that the ordinary HVAC systems of buildings are not optimally configured to handle the concentrated head generated with data centers

Methodology Applied

Scientific EffectConvection: Convection

Data Source

PatentUS8782234B2Arrangement for managing data center operations to increase cooling efficiency

Publication Date: 2014.07.15 SIEMENS INDUSTRY INC

AI summary

A method includes a step of obtaining efficiency characteristic information for each of a plurality of air conditioning units in a location containing a plurality of server computers. The method also includes employing one or more processing units to allocate one or more processing tasks to one of the plurality of server computers based on the efficiency characteristic information.