3D NAND Search Engine With Thermometer Encoding for MLC Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional TCAM technology for big data and AI processing has low memory density and high power usage, while nonvolatile-implemented TCAM techniques require paired memory cells, limiting efficiency in searching and data storage.
Innovation Solution
A 3D search engine architecture using multi-level thermometer encoding for multi-bit cell NAND memory, enabling high performance searching, comparing, and sorting by reducing the number of devices in series and utilizing series-connected floating-gate transistors to store and search quantized data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional TCAM technology is used for searching, then searching functionality is provided, but memory density is low and power usage is high
Solution Approach 1:
The patent replaces conventional TCAM search mechanisms with NAND flash memory-based in-memory searching. Instead of using specialized TCAM circuits, the invention utilizes the existing NAND flash memory structure with modified search operations, substituting a mature, high-density storage technology for a specialized search technology. This achieves both high memory density and low power consumption while maintaining search functionality.
Solution Approach 2:
The patent enables NAND flash memory to perform multiple functions: data storage and in-memory searching. By programming search patterns into the NAND flash memory and utilizing its natural read operations for pattern matching, the memory device serves dual purposes. This eliminates the need for separate TCAM structures and achieves high density with low power usage by leveraging the universal capabilities of NAND flash.
2Reliability
If nonvolatile-implemented TCAM techniques are used, then nonvolatile storage is achieved, but paired memory cells are required which limits efficiency
Solution Approach 1:
The patent extracts the search functionality from the storage structure by using dedicated search circuits (sense amplifiers, decoders) working in conjunction with NAND flash memory. Instead of requiring paired memory cells within the storage array itself, the search operations are performed by separate circuitry that reads from and compares data in the NAND flash. This separation allows efficient searching without compromising storage density or requiring complex paired cell structures.
3Productivity
If higher performance searching is achieved through additional techniques, then searching capability is improved, but device complexity increases
Solution Approach 1:
The patent performs preliminary actions by pre-programming search patterns into the NAND flash memory before the actual search operation. The memory is prepared with training data or pattern templates that enable subsequent rapid matching operations. This preliminary setup allows the search engine to achieve high performance through simple, repetitive read and compare operations rather than complex real-time processing, thereby improving searching capability without significantly increasing device complexity.
Data Source
AI summary
A 3D search engine receives searches for application to word lines of a nonvolatile memory array. Features are stored and searched in accordance with a multi-level thermometer encoding enabling use of multi-bit cell NAND technology, such as MLC, TLC, and QLC. The engine uses two word lines per digit of information of the searches and two memory devices per digit of stored feature to search against. The engine uses respective bit lines of the nonvolatile memory array as respective match lines for searching. Respective memory strings (e.g., NAND memory strings) of the nonvolatile memory array are usable to store respective data words, e.g., corresponding to features to search for. Respective pluralities of the memory strings are coupled in parallel to respective shared bit lines. The engine determines matches according to degree of similarity, such as high to low. The engine is applicable to searching, comparing, and sorting.