Counterflow Pipeline Data Comparison for Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for comparing data elements in two-way pipelines moving in opposite directions face issues with correct data element comparison, leading to inefficiencies and dropped data due to irregular scheduling and synchronization problems, particularly when both pipelines operate simultaneously.
Innovation Solution
The implementation of a counterflow pipeline configuration with comparison units at each node to ensure accurate and quick comparison of data elements moving in opposite directions, using a reference unit to compare data across nodes and adjust storage numbers dynamically to prevent data element omissions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data elements are moved at an irregular schedule to suppress comparison problems in two-way pipelines, then data comparison reliability is improved, but processing efficiency and throughput deteriorate due to frequent stoppages and irregular operations
Solution Approach 1:
The patent applies dynamics by making the pipeline operation mode adjustable between regular and irregular schedules. The control unit dynamically switches between synchronous operation (regular schedule) and asynchronous operation (irregular schedule), allowing the system to adapt to different operational requirements and resolve the contradiction between reliability and efficiency
Solution Approach 2:
The patent changes the operational parameters of the pipeline by introducing adjustable scheduling modes. By modifying the timing and sequence parameters of data element movement, the system can operate in regular intervals for efficiency or irregular intervals for reliability, depending on the comparison requirements
2Speed
If both pipelines operate simultaneously with regular scheduling, then processing speed is improved, but data element comparison accuracy deteriorates due to synchronization problems and interlocking issues
Solution Approach 1:
The patent introduces a control unit as an intermediary that manages the interaction between the two pipelines. This mediator coordinates data element movement and comparison timing, ensuring that even when both pipelines operate simultaneously, the comparison operations maintain accuracy by properly synchronizing the interaction between pipelines
Solution Approach 2:
The pipeline stages perform self-comparison of data elements as they pass through, with each stage independently comparing its local data elements. This self-service mechanism ensures comparison accuracy is maintained at each stage without requiring complex external coordination, allowing simultaneous operation of both pipelines
3Reliability
If a handshake protocol is used for inter-stage communication in synchronous pipelines, then data transmission control is improved, but processing efficiency deteriorates due to bubble formation and transmission delays
Solution Approach 1:
The patent eliminates the handshake protocol's interruptible nature by implementing continuous data flow through the pipeline. Data elements move continuously from stage to stage without being blocked by handshaking delays, maintaining uninterrupted processing while still ensuring proper data transmission through the comparison mechanism at each stage
Data Source
AI summary
An information processing apparatus includes: a first pipeline having first nodes, and moving data held in each first node to a first node located in a first direction; a second pipeline having second nodes corresponding to the first nodes, and moving data held in each second node to a second node located in a second direction that is opposite to the first direction; a first comparison unit arranged to compare data held in a node of interest with data held in a second node corresponding to the node of interest, where the node of interest is one of the first nodes; and a second comparison unit arranged to compare the data held in the node of interest with data held in a second node located one node on an upstream or downstream side of the second node corresponding to the node of interest.