Predictive Server Load Distribution for Data Center Thermal Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data centers face significant energy inefficiencies due to the rapid growth of energy dissipation in servers and cooling systems, with traditional solutions being primarily reactive and failing to optimize energy use across computing and cooling resources.
Innovation Solution
A holistic approach is taken by treating energy as a first-class resource, using predictive thermal models and global schedulers to allocate energy budgets to servers and match cooling efforts with thermal trends, while employing proactive and reactive control mechanisms to manage energy efficiency and cost-effectiveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If servers are overprovisioned with high capacity, then peak performance is improved, but energy efficiency deteriorates due to operating far from peak loading levels
Solution Approach 1:
The patent implements dynamic workload scheduling that continuously monitors and adjusts task allocation across servers based on real-time conditions. The global scheduler dynamically redistributes workloads to optimize energy efficiency while maintaining peak performance capability, allowing servers to operate at optimal loading levels rather than static overprovisioned states
Solution Approach 2:
The system changes operational parameters by adjusting server loading levels and task allocation dynamically. By modifying workload distribution parameters in real-time, the system achieves high energy efficiency at partial loading while preserving the ability to reach peak performance when needed, rather than operating continuously at fixed high-capacity settings
2Temperature
If cooling capacity is increased to handle peak loads, then thermal management is improved, but energy consumption deteriorates due to operating cooling systems at low efficiency points
Solution Approach 1:
The patent employs predictive thermal models that forecast future thermal conditions based on scheduled workloads. This preliminary action allows the cooling system to be proactively adjusted before thermal peaks occur, enabling operation at optimal efficiency points rather than reacting to temperature emergencies that would require excessive cooling capacity
Solution Approach 2:
The system implements feedback mechanisms where thermal models continuously monitor actual temperature conditions and adjust cooling schedules accordingly. This closed-loop control ensures cooling systems operate at efficient loading levels by matching cooling capacity to actual thermal demands rather than maintaining fixed high-capacity operation
3Ease of operation
If reactive control mechanisms are used to manage energy, then responsiveness to current conditions is improved, but overall energy efficiency deteriorates due to lack of proactive optimization
Solution Approach 1:
The patent combines reactive monitoring with proactive optimization by using predictive thermal models to forecast future conditions and pre-adjust workload schedules and cooling systems. This preliminary action maintains responsiveness to current conditions while adding forward-looking optimization that improves overall energy efficiency
Solution Approach 2:
The system introduces predictive thermal models as an intermediary between reactive sensors and control actions. These models process current sensor data and generate forecasts that mediate the control decisions, enabling both responsive reaction to current conditions and proactive optimization based on predicted future states
4Area of stationary object
If servers are packed densely in data centers, then space utilization is improved, but energy dissipation worsens due to concentrated heat generation
Solution Approach 1:
The patent applies local quality by implementing rack-specific and server-specific workload scheduling and cooling strategies. Instead of uniform treatment, the system tailors task allocation and cooling efforts to local thermal conditions and server capabilities, enabling dense packing while managing concentrated heat generation through localized optimization
Solution Approach 2:
The system changes operational parameters by dynamically adjusting workload distribution and cooling settings at local levels. By modifying loading levels and cooling capacity parameters individually for different racks and servers, the system maintains high space utilization while managing energy dissipation through localized parameter optimization rather than uniform operation
Data Source
AI summary
A method for controlling a data center, comprising a plurality of server systems, each associated with a cooling system and a thermal constraint, comprising: a concurrent physical condition of a first server system; predicting a future physical condition based on a set of future states of the first server system; dynamically controlling the cooling system in response to at least the input and the predicted future physical condition, to selectively cool the first server system sufficient to meet the predetermined thermal constraint; and controlling an allocation of tasks between the plurality of server systems to selectively load the first server system within the predetermined thermal constraint and selectively idle a second server system, wherein the idle second server system can be recruited to accept tasks when allocated to it, and wherein the cooling system associated with the idle second server system is selectively operated in a low power consumption state.