Thread Scheduling with Temporal Core Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional thread scheduling in multi-processor systems leads to uneven wear across processor cores, resulting in reduced performance, reliability, and efficiency as some cores are used more frequently than others, causing premature degradation and overheating.
Innovation Solution
A method that schedules threads based on the historical active time of processor cores, prioritizing the least used cores to distribute workload evenly, thereby preventing any single core from wearing out quickly and ensuring uniform operation within a safe thermal range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional thread scheduling is used, then system throughput is maintained, but processor wear becomes uneven and reliability decreases
Solution Approach 1:
The system performs preliminary tracking of active time for each processor core before scheduling threads. By maintaining historical usage data and identifying the least used core in advance, the scheduler can make informed decisions that distribute wear evenly across cores, preventing any single core from degrading prematurely while maintaining overall system throughput.
2Temperature
If threads are scheduled on available processors, then processing speed is maintained, but thermal range safety is compromised
Solution Approach 1:
The system implements feedback by continuously monitoring the active time and thermal state of processor cores. This feedback information is used to dynamically adjust scheduling decisions, selecting cores that are currently within safe thermal ranges. The feedback loop ensures that processing continues at high speed while respecting thermal constraints, as the scheduler adapts to real-time conditions.
3Productivity
If frequent thread migration occurs, then load balancing is improved, but performance degrades due to frequency ramp time
Solution Approach 1:
The system applies local quality by making scheduling decisions based on the specific characteristics of each processor core, particularly its historical active time and current thermal state. Rather than applying a uniform scheduling policy, the system tailors thread assignment to the local condition of each core, selecting the most appropriate target core for each thread while minimizing unnecessary migrations and associated frequency ramp time.
Data Source
AI summary
Systems, apparatuses and methods may provide for technology that identifies a thread and selects a core from a plurality of processor cores in response to the selected core being available while satisfying a least used condition with respect to the plurality of processor cores. The technology may also schedule the thread to be executed on the selected core.