Parallel Processing Control for Low-Overhead Arithmetic Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As the number of parallel processing cores increases, the allocation of processing tasks becomes complex, leading to concerns about increased overhead in automated driving systems that rely on CPUs, PLDs, and GPUs for high-speed arithmetic operations.
Innovation Solution
An electronic control device with two processing control units that alternately start arithmetic processing, where the second processing unit begins after the first has started but before it finishes, to manage and allocate tasks efficiently, thereby reducing overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the number of parallel processing cores is increased to accelerate processing, then productivity is improved, but device complexity and overhead increase
Solution Approach 1:
The patent segments processing into distinct phases (data acquisition, preprocessing, feature extraction, recognition, control) and assigns different cores to different phases. This segmentation allows each core to specialize in specific tasks, improving processing speed while keeping the allocation scheme manageable rather than requiring complex dynamic scheduling across all tasks simultaneously
Solution Approach 2:
The patent establishes processing allocation in advance based on task characteristics before execution begins. The allocation scheme is predetermined according to the nature of each processing task, eliminating the need for complex real-time allocation decisions during parallel execution, thus reducing overhead while maintaining high productivity
2Productivity
If the number of parallel processing cores is increased to accelerate processing, then productivity is improved, but overhead increases
Solution Approach 1:
By segmenting processing into distinct phases and assigning dedicated cores to each phase, the patent eliminates the overhead of dynamic task allocation and context switching that would occur with more flexible scheduling approaches. Each core continuously executes its assigned phase without interruption, reducing energy consumption from allocation operations while maintaining high processing throughput
Solution Approach 2:
Each processing core independently executes its assigned tasks without requiring centralized coordination or communication for allocation decisions. The cores operate autonomously based on predetermined allocation rules, eliminating the overhead of inter-core communication and centralized scheduling, thus reducing energy loss while achieving parallel processing acceleration
Data Source
AI summary
An electronic control device includes a first processing control unit and a second processing control unit, wherein: the first processing control unit and the second processing control unit alternately start arithmetic processing with information of an external environment as an input; and second arithmetic processing of the second processing control unit is started after first arithmetic processing by the first processing control unit is started and before the first arithmetic processing is ended.