Storage Device Stream Filtering for Host Load Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor memory devices, particularly flash memory-based storage devices, face challenges in efficiently processing and filtering data without overburdening the host device, leading to increased power consumption and data transfer inefficiencies.
Innovation Solution
A stream filtering method and device architecture that employs a filter manager to control multiple stream filters and a merge filter, allowing parallel data processing and filtering within the storage device, reducing the host's data processing burden and minimizing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If data processing and filtering are performed by the host device, then processing capability is sufficient, but power consumption increases and data transfer efficiency decreases
Solution Approach 1:
The patent extracts the data filtering function from the host device and implements it within the storage device itself through dedicated filter hardware (first filter, second filter, merge filter). This extraction reduces the data processing burden on the host, thereby lowering power consumption while improving data transfer efficiency by processing data closer to its source.
Solution Approach 2:
The patent introduces an intermediary filtering system within the storage device that acts as a mediator between the raw data storage and the host device. The filter manager and filter hardware process data locally, serving as an intermediate layer that reduces unnecessary data transfers to the host while maintaining processing efficiency.
2Ease of operation
If data filtering is performed within the storage device, then host processing burden is reduced, but device complexity increases
Solution Approach 1:
The patent segments the filtering function into distinct modular components: a filter manager for control, a first filter for initial filtering, a second filter for additional filtering, and a merge filter for combining results. This segmentation allows the complex filtering operation to be divided into manageable, independent modules that can be controlled and maintained separately, reducing the operational burden on the host.
Solution Approach 2:
The filter manager acts as an intermediary control unit that manages the complex filtering operations internally within the storage device. By localizing this control mechanism, the device handles its own complexity internally while presenting a simplified interface to the host, thereby reducing the host's processing burden without exposing the full complexity to the external system.
3Speed
If multiple filters process data in parallel, then filtering speed improves, but device complexity increases
Solution Approach 1:
The patent divides the filtering process into multiple parallel filter units (first filter, second filter, merge filter) that operate simultaneously on different aspects of the data. Each filter handles specific filtering tasks independently, enabling parallel processing that increases filtering speed while keeping each individual filter unit relatively simple in structure.
Solution Approach 2:
The patent combines multiple parallel filter operations through a merge filter that integrates the results from the first and second filters. This merging approach allows the system to maintain parallel processing capabilities for speed while consolidating the complex operations into a unified output structure, balancing speed improvement with manageable complexity.
Data Source
AI summary
A storage device may include a main storage part including one or more memories; and a controller configured to control an overall operation of the main storage part. The controller includes a filter manager configured to store data format information and a filtering condition provided from a host; one or more stream filters configured to search and project data stored in the one or more memories in parallel in response to a control of the filter manager to produce searched and projected data; and a merge filter configured to merge the searched and projected data of the one or more stream filters in response to the control of the filter manager.