SSD Controller Self-Reporting for Real-Time Performance Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Host software lacks real-time information on non-volatile memory (NVM) storage device performance capabilities, leading to unreliable performance planning and inefficient resource allocation due to varying SSD performance levels and degradation over time.
Innovation Solution
Non-volatile memory storage devices equipped with a device controller that tracks and reports performance capabilities to host software, including changes through asynchronous events, allowing for dynamic resource allocation and optimal application performance mapping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If host software uses datasheet-based performance information for capacity planning, then performance planning can be performed, but the information becomes unreliable as SSDs wear out and performance degrades over time
Solution Approach 1:
The SSD controller continuously monitors actual performance metrics (IOPS, throughput, latency) and provides feedback to the host software through identify commands. This real-time feedback mechanism allows the host to track performance degradation and adjust capacity planning dynamically, resolving the unreliability of static datasheet information.
Solution Approach 2:
The SSD controller autonomously measures and reports its own performance capabilities without requiring external testing equipment or manual intervention. The controller self-monitors performance metrics and provides this information to the host software, eliminating the need for the host to maintain external performance databases.
2Adaptability or versatility
If host software maintains a database of performance capabilities for multiple SSD models, then performance mapping can be performed, but the system complexity increases and database updates are required when configuration changes
Solution Approach 1:
The performance capability information is extracted from the host software's internal database and relocated to the SSD controller itself. The controller now stores and provides performance metrics directly to the host through standardized commands, eliminating the need for the host to maintain complex performance databases.
Solution Approach 2:
The identify commands in the NVMe interface are enhanced to provide multiple functions: device identification, capacity information, and performance capability reporting. This multi-functional approach consolidates what would otherwise require separate database lookups and configuration management.
3Quantity of substance
If aggregate performance data is used for SSDs with multiple namespaces, then device-level performance can be reported, but it becomes less useful for mapping application-level performance requirements to specific namespaces
Solution Approach 1:
Performance capability reporting is segmented by namespace, with the controller providing separate performance metrics for each namespace. This allows the host software to map application performance requirements to specific namespaces with precise granularity, rather than using aggregate device-level data.
Data Source
AI summary
Provided is a storage device including a first non-volatile storage media having first performance capabilities; a second non-volatile storage media having second performance capabilities different from the first performance capabilities; and a device controller configured to report to a host software the first performance capabilities, the second performance capabilities, changes to the first performance capabilities, and changes to the second performance capabilities.