On-Chip NVM Search Circuitry for Low-Bandwidth Data Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data storage systems, such as SSDs with NAND flash memory, face inefficiencies when searching for data within a range of logical block addresses (LBAs), as they often fetch and discard large amounts of unnecessary data, leading to increased processing time and reduced endurance of non-volatile memory dies.
Innovation Solution
Implementing on-chip search circuitry within the non-volatile memory (NVM) die to perform searches directly, using scrambling patterns and dedicated search circuitry to identify and return only relevant data, thereby reducing the need for host device processing and minimizing read cycles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the host fetches all data from NVM corresponding to an LBA range for searching, then the search can be performed, but considerable unnecessary data is transferred and processed, increasing power consumption and bus bandwidth usage
Solution Approach 1:
The patent extracts only the relevant search results from the NVM array and returns them to the host, rather than transferring all data in the LBA range. The search circuitry within the NVM die identifies and outputs only matching data, eliminating the need for the host to process and discard unnecessary data.
Solution Approach 2:
The patent introduces search circuitry as an intermediary component within the NVM die that performs the search operation before data transfer. This intermediary processing stage filters the data locally, so only relevant results traverse the bus to the host, reducing overall system power consumption.
2Productivity
If the host fetches all data from NVM corresponding to an LBA range for searching, then the search can be performed, but considerable unnecessary data is transferred, increasing bus bandwidth usage
Solution Approach 1:
The search circuitry extracts and returns only the specific data items that match the search criteria from the NVM array, rather than transferring the entire LBA range data to the host. This extraction occurs at the source, minimizing the quantity of data that must be transferred over the bus.
3Productivity
If the host fetches all data from NVM corresponding to an LBA range for searching, then the search can be performed, but processing time increases due to unnecessary data handling
Solution Approach 1:
The search operation is performed preliminarily within the NVM die before data transfer to the host. The search circuitry identifies matching data items in advance, so the host receives only pre-filtered results, eliminating the time the host would spend processing and discarding non-matching data.
4Productivity
If read cycles are increased to search for data, then data can be retrieved, but the endurance of NVM dies is reduced
Solution Approach 1:
The NVM die performs the search operation itself using integrated search circuitry, eliminating the need for the host to read all data and perform the search externally. This self-service capability reduces the number of read cycles required, as the search is accomplished through specialized circuitry rather than repeated read operations.
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.