Storage Engine Power Estimation Under a Strict Power Budget
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage devices face challenges in efficiently managing power consumption, leading to sub-optimal performance and potential failure due to deviations from a strict power budget, which current characterization procedures are lengthy and inflexible.
Innovation Solution
Implementing a power sensor to measure total power consumption, determining power-per-processing event values using a least-squares procedure, and controlling power delivery based on these values to adaptively manage power within the device's budget.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional power management methods are used, then device simplicity is maintained, but power consumption control precision deteriorates leading to sub-optimal performance
Solution Approach 1:
The patent replaces traditional mechanical power management approaches with a data-driven statistical model using least-squares estimation. The system substitutes complex hardware-based power control mechanisms with a mathematical framework that correlates processor activity states with power consumption measurements, achieving precise power control through computational methods rather than mechanical or electronic control circuits.
Solution Approach 2:
The patent introduces an intermediary statistical model that acts as a mediator between processor activity monitoring and power consumption control. Instead of directly controlling power based on raw measurements, the system uses least-squares estimation to create a predictive relationship between processor states and power usage, allowing for optimized power management decisions based on estimated rather than directly measured values.
2Measurement precision
If detailed power characterization procedures are implemented, then power measurement accuracy is improved, but characterization time increases becoming lengthy and inflexible
Solution Approach 1:
The patent performs preliminary least-squares estimation during initial system operation to establish the relationship between processor activity states and power consumption. By pre-characterizing the power model during normal operation rather than requiring separate lengthy characterization procedures, the system captures power consumption patterns without stopping or slowing down device operation, thus eliminating time losses associated with traditional characterization methods.
Solution Approach 2:
The patent enables continuous power characterization and model updating during normal device operation. The least-squares estimation process operates continuously in the background, continuously refining the power model without interrupting processor operations or requiring dedicated characterization time windows. This maintains useful computational action throughout the characterization process rather than pausing for separate measurement phases.
3Reliability
If fixed power budgets are enforced, then power consumption is controlled, but adaptability to device changes deteriorates
Solution Approach 1:
The patent implements a dynamic power management system where the least-squares estimation model is continuously updated based on changing device conditions, workload patterns, and power consumption characteristics. Rather than using a static fixed power budget, the system adapts the power model parameters in real-time to reflect current device state, maintaining power budget compliance while automatically adjusting to device changes, aging, temperature variations, and workload evolution.
Solution Approach 2:
The patent incorporates feedback mechanisms where actual power consumption measurements are continuously compared with least-squares estimated values, and the model is refined based on the difference between predicted and actual power usage. This closed-loop feedback ensures power budget compliance while automatically adapting the power model to reflect real device behavior, capturing changes in device characteristics over time without requiring manual reconfiguration.
Data Source
AI summary
Methods and apparatus for power management in data storage devices are provided. One such data storage device (DSD) includes a non-volatile memory (NVM), a set of hardware processing engines, and a power sensor to measure a total power consumption of the set of hardware processing engines. A processor is configured to determine a power-per-processing event value for each of the set of processing engines based on total power consumption measurements, then control delivery of power to the processing engines based on the power-per-processing event values in accordance with a power budget. In some examples, the DSD employs a least-squares procedure to estimate the power-per-processing event values so the values can be determined without needing to measure the individual power consumption of the processing engines. Procedures are also provided for assessing the accuracy of the power-per-processing event values and for controlling further operations based on the assessment.