User-Perceivable Latency Metric for Mobile Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing performance metrics fail to reliably predict whether low-power storage units, such as eMMC SSDs, integrated into mobile devices will provide a satisfying user experience, as they may not meet performance targets for mobile workloads despite meeting conventional performance standards.
Innovation Solution
A new metric, user-perceivable latency, is introduced to characterize device performance by measuring and rating latencies based on their impact on user experience, focusing on durations and attributes of operations to differentiate between perceivable and non-perceivable delays, thereby guiding component selection and device design to meet user expectations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional performance metrics are used to evaluate storage devices, then manufacturing and testing are simplified, but the metrics fail to accurately predict user-perceivable performance
Solution Approach 1:
The patent transforms the performance measurement approach by changing the parameters being measured from conventional metrics (transfer rates, IOPS) to user-perceivable latency metrics. This involves measuring latency at multiple points in the data path and applying different weights to different latency components based on their impact on user experience. The measurement system evaluates latency contributions from various device components and combines them with weighting factors that reflect actual user perception, thereby achieving more accurate performance prediction without requiring complex hardware modifications.
2Use of energy by moving object
If low-power storage components are selected to meet design constraints, then energy consumption is reduced, but performance may fall below user satisfaction thresholds
Solution Approach 1:
The patent enables better component selection by changing the performance evaluation parameters from conventional metrics to user-perceivable latency metrics. This allows designers to accurately assess whether low-power components will meet user satisfaction thresholds before finalizing design decisions. The measurement system provides detailed latency breakdowns that show which components contribute most to perceived delays, enabling optimized component selection that balances power consumption with actual user experience.
Solution Approach 2:
The patent implements a feedback mechanism where latency measurements from actual device operation are used to refine component selection and design decisions. The system continuously monitors and evaluates latency contributions from different components, providing feedback that guides design optimizations. This feedback loop allows designers to iteratively improve the balance between power consumption and user-perceived performance by identifying specific components that contribute most to latency.
3Loss of information
If comprehensive latency measurements are taken for all operations, then complete performance data is obtained, but processing and analysis become computationally intensive
Solution Approach 1:
The patent extracts only the most relevant latency information from comprehensive measurements by identifying and isolating user-perceivable latency components. Rather than processing all raw latency data, the system extracts specific latency measurements from critical operation points in the data path and applies weighting factors to these extracted values. This extraction approach maintains data completeness for analysis purposes while significantly reducing the computational burden of generating performance ratings.
Solution Approach 2:
The patent applies different processing quality levels to different latency measurements based on their importance to user experience. Critical latency components that directly impact user perception are measured and processed with high precision and appropriate weighting, while less significant latency components are processed with reduced detail. This local quality approach ensures that the most important performance information is preserved and accurately evaluated while reducing overall computational requirements.
Data Source
AI summary
A method and an apparatus for characterizing performance of a device based on user-perceivable latency. To characterize device performance, a value of a metric may be computed from latencies of operations performed by the device. In computing a value of a metric, latencies may be treated differently, such that some latencies perceivable by a user of the device may have a greater impact on the value of the metric than other latencies that either are not perceivable or are perceived by the user to a lesser degree. Such a performance metric based on user-perceivable latency facilitates identification of computing device that provide a desirable user experience.