Dynamic Data Center Cooling via Variable Liquid Flow
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data centers face significant challenges in managing power consumption and heat removal, as high power usage by microprocessors generates substantial heat, leading to increased cooling costs and potential microprocessor errors and failures. Existing cooling systems often lack the necessary capacity to efficiently handle peak loads, resulting in inefficient energy use and potential overheating.
Innovation Solution
The proposed solution involves circulating a cooling airflow to prioritize high-power usage areas while adjusting the flow rate of a cooling liquid to maintain constant heat removal rates. This is achieved by monitoring power usage and temperature levels, rerouting cooling liquids, and modulating airflow to ensure efficient heat dissipation across the data center, even when cooling capacity is limited.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Temperature
If cooling capacity is increased to handle peak loads, then heat removal capability is improved, but energy consumption and system cost increase
Solution Approach 1:
The cooling system dynamically adjusts cooling liquid flow rates based on real-time monitoring of computing device power usage. The system transitions from static cooling capacity to dynamic adjustment, increasing flow to high-power devices and decreasing flow to low-power devices, thereby matching cooling output to actual heat generation and reducing wasted energy.
Solution Approach 2:
The system applies different cooling strategies to different portions of computing devices based on their individual power usage characteristics. High-power usage portions receive increased cooling liquid flow while low-power usage portions receive decreased flow, creating localized quality differences in cooling intensity that optimize overall system efficiency.
2Use of energy by moving object
If cooling capacity is reduced to lower costs, then energy consumption decreases, but heat removal capability becomes insufficient during peak loads
Solution Approach 1:
The system continuously monitors power usage of computing devices and uses this feedback to adjust cooling liquid flow rates in real-time. This closed-loop control ensures that cooling capacity is automatically increased when heat generation increases and decreased when heat generation decreases, maintaining cooling adequacy while optimizing energy consumption.
Solution Approach 2:
The system proactively adjusts cooling liquid flow rates based on monitored power usage trends before overheating occurs. By detecting increased power usage early and preemptively increasing cooling flow to affected devices, the system prevents temperature excursions that could compromise reliability.
3Device complexity
If uniform cooling is applied to all computing devices, then system simplicity is maintained, but cooling efficiency decreases when power usage varies
Solution Approach 1:
The system transitions from uniform cooling to localized quality-based cooling by monitoring individual device power usage and adjusting cooling liquid flow rates accordingly. Devices operating at high power levels receive increased cooling flow while devices at low power levels receive decreased flow, optimizing cooling efficiency without requiring complete system redesign.
Solution Approach 2:
The system applies partial cooling action to different device portions based on their power usage needs. Rather than applying full cooling capacity uniformly, the system provides only the necessary cooling amount to each device portion, avoiding excessive cooling to low-power devices and reducing overall energy consumption.
4Temperature
If cooling liquid flow rate is increased to all devices, then heat removal is improved, but energy consumption and pump workload increase
Solution Approach 1:
The system changes the flow rate parameter of cooling liquid dynamically based on monitored power usage conditions. By adjusting flow rates from a constant high level to variable levels matched to actual cooling needs, the system maintains effective heat removal while significantly reducing pump energy consumption during periods when full cooling capacity is not required.
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 allows for more efficient heat removal, reducing energy consumption and extending the lifespan of microprocessors by maintaining optimal temperatures, even during peak usage periods, while also enabling the construction of data centers with reduced cooling capacity that can still meet year-round cooling needs.
Implementation Method 1
circulating a cooling liquid to cool a second portion of the plurality of computing devices
Implementation Method 2
circulating a cooling airflow to cool a plurality of computing devices supported in racks in a data center
Data Source
AI summary
Techniques for cooling a data center include circulating a first cooling medium to cool a plurality of rack-mounted computers; circulating a second cooling medium to cool the plurality of rack-mounted computers; determining that a first portion of the plurality of rack-mounted computers is operating at a power usage above a threshold power usage; adjusting at least one of a flow rate of the first or second cooling mediums to cool a second portion of the plurality of rack-mounted computers; and rerouting a portion at least one the first or second cooling mediums to cool the first portion of the plurality of rack-mounted computers.