LUT-Based TCAM Architecture for Single-Cycle Search and Update
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
TCAM architectures face challenges with low storage density, high power consumption, and high cost per bit, while also experiencing longer update latencies and suspended search operations during updates, which hinder high-speed and efficient search and update processes.
Innovation Solution
A TCAM architecture utilizing two memory elements per cell, where the search key is used as an address, eliminating the need for comparison between the search key and stored data, and an FPGA-based design with look-up tables that allow direct access and simultaneous update of bits, enabling high-speed single-cycle searches and updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If traditional TCAM comparison logic is used, then search functionality is achieved, but power consumption increases and storage density decreases
Solution Approach 1:
The patent extracts and removes the comparison logic from the TCAM cell structure, retaining only the storage function. By taking out the energy-intensive comparison circuits, the design achieves lower power consumption while maintaining search capability through a different mechanism (using the stored value directly as the match result).
Solution Approach 2:
The patent uses the stored data value itself as the match indicator, effectively copying the storage function to serve the search function. Instead of comparing search key with stored data, the stored data is directly used to determine match status, eliminating the need for separate comparison circuits.
2Productivity
If TCAM update operations are performed, then data can be refreshed, but search operations must be suspended and latency increases
Solution Approach 1:
The patent segments the TCAM cell into separate storage elements independent of comparison logic. This segmentation allows update operations to be performed on storage elements without affecting the search pathway, enabling concurrent search and update operations that eliminate search latency during updates.
Solution Approach 2:
The patent prepares data in advance by storing it in a format that can serve both as stored information and as match result indicators. This preliminary preparation of data in dual-purpose storage elements allows immediate response to search operations without requiring additional comparison steps that would cause latency.
3Quantity of substance
If conventional TCAM cells with comparison logic are used, then match detection is possible, but storage density decreases
Solution Approach 1:
The patent makes the storage element universal by having it perform both storage and match detection functions. The same storage element that holds the data also provides the match result, eliminating the need for separate comparison circuits and thereby increasing storage density while maintaining accurate match detection capability.
Solution Approach 2:
Instead of using comparison logic to determine matches, the patent inverts the approach by using the presence and value of stored data itself as the match indicator. This inversion eliminates comparison circuits and increases storage density while preserving match detection accuracy through the direct use of stored values.
Data Source
AI summary
According to an embodiment of the present invention, there is provided a TCAM architecture in which a content-based search is conductible in such a manner that a search key to be searched for is used as an address of a memory element that makes up a TCAM cell and that an output from the memory element reflects whether a match or a mismatch is found as a result of the search. The memory element may be a look-up table.