Intelligent Thermal Throttling for Flash Storage Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash storage systems face significant thermal constraints, leading to inefficient performance throttling due to static throttling methods that introduce fixed delays based on pre-defined temperature thresholds, which can result in performance wastage as they accommodate worst-case scenarios rather than actual temperature conditions.
Innovation Solution
A data-driven intelligent thermal throttling method is implemented, where a controller estimates a future temperature curve using a saturating temperature (K) and curvature factor (R), dynamically adjusting sampling windows, frequency, and number of samples to determine optimal throttling delays, thereby providing adaptive performance adjustments based on predicted temperature changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If static throttling with fixed delays is used based on pre-defined temperature thresholds, then thermal constraints are managed, but performance is wasted due to accommodating worst-case scenarios rather than actual temperature conditions
Solution Approach 1:
The patent transforms static throttling into dynamic throttling by continuously monitoring temperature and adjusting throttling delays in real-time based on actual temperature conditions. The system adapts the throttling strategy dynamically rather than using fixed pre-defined thresholds, allowing performance to be optimized according to actual thermal states rather than worst-case scenarios.
Solution Approach 2:
The patent changes the throttling delay parameter dynamically based on temperature measurements. Instead of using fixed delays, the system adjusts the magnitude of throttling delays according to actual temperature conditions, thereby optimizing performance while still managing thermal constraints effectively.
2Temperature
If throttling is applied to mitigate temperature issues, then thermal constraints are addressed, but performance is reduced
Solution Approach 1:
The patent implements a feedback mechanism where temperature is continuously monitored and the throttling strategy is adjusted based on actual temperature readings. This closed-loop control allows the system to apply throttling only when necessary and at the appropriate magnitude, minimizing performance impact while effectively managing thermal constraints.
Solution Approach 2:
The system dynamically adjusts throttling intensity based on real-time temperature conditions rather than applying fixed throttling. This allows the system to maintain high performance when temperatures are acceptable and apply throttling only when thermal constraints are approached, optimizing the balance between temperature management and performance.
3Device complexity
If fixed delays are introduced at temperature thresholds, then thermal management is simplified, but performance wastage occurs due to over-throttling
Solution Approach 1:
The patent replaces simple fixed-delay throttling with a dynamic throttling mechanism that continuously adapts to temperature conditions. While this increases control complexity compared to fixed thresholds, it eliminates performance wastage by applying throttling only when and to the extent actually needed based on real-time temperature measurements.
Data Source
AI summary
A storage system and method for data-driven intelligent thermal throttling are provided. In one embodiment, the storage system comprises a memory and a controller. The controller is configured to determine a temperature of the memory, estimate a future temperature curve based on the temperature of the memory, and determine a memory throttling delay to apply based on the estimated future temperature curve. Other embodiments are provided.