Pattern Recognition Processor State Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Pattern-recognition processing in computing is bottlenecked by the need to search large volumes of data for multiple patterns, leading to increased delays as the number of patterns increases, and the time-consuming process of saving and restoring state information in processors.
Innovation Solution
A pattern-recognition processor that uses parallel finite state machines to evaluate multiple search criteria simultaneously, with a state variable storage array for efficient saving and restoring of states, allowing it to maintain performance regardless of the number of search criteria.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sequential pattern searching is used to identify multiple patterns in a data stream, then the system can detect various spam and malware patterns, but the processing delay increases with the number of patterns
Solution Approach 1:
The pattern recognition processor is divided into multiple parallel finite state machines, each capable of independently searching for specific patterns in the data stream. This segmentation allows simultaneous pattern detection without sequential delays, resolving the contradiction between comprehensive pattern detection and processing speed.
Solution Approach 2:
The system transitions from sequential (one-dimensional) pattern searching to parallel (multi-dimensional) processing by implementing multiple finite state machines that operate simultaneously. This dimensional change enables the system to maintain constant processing delay regardless of the number of patterns being searched.
2Adaptability or versatility
If the processor switches between multiple data streams and tasks, then it can handle diverse processing requirements, but saving and restoring state information takes significant time
Solution Approach 1:
The system maintains copies of state information in dedicated state variable storage arrays that are directly accessible to each finite state machine. When switching between data streams or tasks, the processor can rapidly copy or switch between state copies without time-consuming save/restore operations, enabling fast context switching while maintaining adaptability.
Data Source
AI summary
Systems and methods are disclosed for saving and restoring the search state of a pattern-recognition processor. Embodiments include a pattern-recognition processor having a state variable array and a state variable storage array stored in on-chip memory (on-silicon memory with the processor). State variable storage control logic of the pattern-recognition processor may control the saving of state variables from the state variable array to the state variable storage array. The state variable storage control logic may also control restoring of the state variables from the state variable storage array to restore a search state.