On-Chip NVM Wordline Search for Low-Read Data Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In data storage devices like SSDs with non-volatile memory arrays, searching for specific data within a range of logical block addresses often involves fetching and processing unnecessary data, leading to inefficient processing and reduced endurance of the NVM dies due to excessive read cycles and the need for large index tables.
Innovation Solution
Implementing on-chip search circuitry within the NVM die to directly search for data by sensing wordlines serially and applying voltages, allowing for selective transfer of only matching data to the host, reducing the need for host-based processing and minimizing read operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the host fetches all data from NVM within an LBA range for searching, then the search can be performed, but considerable unnecessary data is transferred and processed, reducing efficiency and increasing power consumption
Solution Approach 1:
The patent extracts only the necessary data (matching search criteria) from the NVM array and transfers it to the host, rather than transferring all data within the LBA range. The search circuitry identifies and outputs only relevant data pages, eliminating the waste of transferring and processing unnecessary data.
Solution Approach 2:
The search operation is performed preliminarily within the NVM die before data transfer to the host. The search circuitry scans and identifies matching data pages in advance, so that only relevant data needs to be transferred to the host, reducing subsequent processing overhead and power consumption.
2Reliability
If the host fetches all data from NVM for searching, then complete search coverage is achieved, but excessive read cycles reduce NVM die endurance
Solution Approach 1:
The search circuitry extracts only the specific data pages that match the search criteria from the NVM array, avoiding repeated read cycles for the same data. By performing the search within the die and outputting only matching pages, the system reduces the total number of read operations required.
Solution Approach 2:
The NVM die performs the search operation itself through integrated search circuitry, rather than requiring the host to read all data and perform searching externally. This self-service capability reduces the burden on the host and minimizes the number of read cycles needed from the NVM array.
3Measurement precision
If all data within LBA range is transferred to host for searching, then accurate search results are obtained, but processing time and host resources are wasted on unnecessary data
Solution Approach 1:
The search circuitry extracts and outputs only data pages that match the search criteria, eliminating the need for the host to process and filter large amounts of irrelevant data. This extraction approach maintains search accuracy while dramatically reducing processing time and host resource utilization.
Solution Approach 2:
The filtering and identification of matching data pages is performed preliminarily within the NVM die before transfer to the host. By pre-processing the data and identifying matches in advance, the system avoids the time-consuming operation of transferring and then filtering large volumes of unnecessary data at the host.
4Measurement precision
If a large index table is maintained for LBA-PBA translation, then data location mapping is accurate, but memory resources are consumed and access complexity increases
Solution Approach 1:
The search operation is segmented into distinct phases: searching within the NVM die using integrated circuitry, identifying matching pages, and transferring only those pages to the host. This segmentation eliminates the need for a large centralized index table, as the search circuitry directly locates matching data pages through localized operations.
Solution Approach 2:
The search circuitry integrated within the NVM die acts as an intermediary between the host and the stored data. Instead of requiring the host to manage large index tables for LBA-PBA translation, the search circuitry performs the mapping and identification functions locally, simplifying the overall system architecture.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances processing efficiency by transferring only relevant data, reduces power consumption, and decreases the load on the NVM dies, thereby improving overall performance and extending their endurance.
Implementation Method 1
sense wordlines of stored data in the NVM array by applying voltages on the wordlines serially
Data Source
AI summary
The disclosure relates in some aspects to on-chip processing circuitry formed within the die of a non-volatile (NVM) array to perform data searches. In some aspects, the die includes components configured to sense wordlines of stored data in the NVM array by applying voltages on the wordlines serially, and then search for an input data pattern within the serially-sensed wordlines. In some examples, the components of the die include latches and circuits configured to perform bitwise latch logic search operations. In other examples, the search components are configured with under-the-array or next-to-the-array dedicated search circuitry that uses registers and/or random access memory (RAM). Other aspects relate to a separate controller device for controlling the on-chip NVM search operations. For example, the controller may determine whether to search for data using search components of the NVM die or processors of the controller based, e.g., on a degree of fragmentation of data.


