Thermal Mitigation in Multi-Core Processors via Load Redistribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-core processors in portable devices face thermal management challenges due to limited power budgets and form factors, leading to performance limitations when temperatures reach mitigation thresholds, and additional cores may not directly enhance performance for single or dual core-intensive applications.
Innovation Solution
A thermal control module in user equipment (UE) monitors core temperatures and transfers loads from overheated cores to cooler cores, adjusting processing frequencies and voltages to maintain temperatures below mitigation thresholds, thereby extending operational time and improving performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the processing frequency and voltage of a core are increased to improve performance, then the processing speed improves, but the temperature of the core increases
Solution Approach 1:
The system dynamically adjusts the processing frequency and voltage of cores based on real-time temperature monitoring. When a core's temperature exceeds a threshold, the system reduces its frequency and voltage to lower temperature, and transfers its workload to other available cores. This dynamic adaptation resolves the contradiction by allowing high performance when temperatures are acceptable and switching to thermal management when temperatures rise.
Solution Approach 2:
The system changes operational parameters (frequency and voltage) of processor cores based on temperature conditions. By monitoring core temperatures and adjusting frequency/voltage accordingly, the system can operate at high performance levels when cool and reduce to lower power states when hot, effectively managing the trade-off between speed and temperature.
2Productivity
If multiple cores are used to process loads simultaneously, then productivity improves, but thermal management becomes more complex
Solution Approach 1:
Each core monitors its own temperature and can independently adjust its frequency and voltage based on its thermal state. The system autonomously redistributes workloads among cores without requiring complex centralized thermal management, allowing multiple cores to operate in parallel while each self-regulates its thermal conditions.
Solution Approach 2:
The system divides the thermal management task into independent per-core monitoring and control units. Each core has its own temperature sensor and control logic, allowing parallel thermal management of multiple cores rather than requiring a single complex centralized system. This segmentation enables scalable multi-core operation with manageable thermal control.
3Temperature
If the processing frequency of a core is reduced to lower temperature, then temperature decreases, but processing speed decreases
Solution Approach 1:
When one core reduces its frequency to lower temperature, the system combines its workload with other available cores. By merging the processing tasks across multiple cores, the system maintains overall processing throughput even though individual core speeds are reduced, resolving the contradiction between temperature reduction and performance maintenance.
Solution Approach 2:
The system dynamically reallocates workloads between cores based on their thermal and performance states. When a core slows down due to thermal constraints, the system dynamically shifts its tasks to other cores that can operate at higher frequencies, maintaining overall system throughput while managing individual core temperatures.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A thermal controller for managing thermal energy of a multi-core processor is provided. The cores include a first core processing a load and remaining cores. The thermal controller is configured to determine that a temperature of the first core is greater than a first threshold, determine a temperature of a second core of the remaining cores in response to determining that the temperature of the first core is greater than the first threshold, and determine whether the temperature of the second core is greater than or less than a second threshold. The thermal controller is configured to transfer at least a portion of the load of the first core to the second core in response to determining that the temperature of the first core is greater than the first threshold and based on whether the temperature of the second core is greater than or less than the second threshold.