Parallel Processing Architecture for Optimized Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computational optimization technologies require significant resources due to large solution spaces and high computational complexity, making it difficult to generate solutions in useful time frames, especially for combinatorial problems with multiple objectives and constraints.
Innovation Solution
Implementing parallel processing architectures, such as GPUs, to perform end-to-end parallel optimization, allocating operations across multiple processing units for diversity management and genetic/evolutionary searching, and using data structures to facilitate efficient parallel processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If exhaustive search is used to ensure complete solution space coverage, then solution completeness is improved, but computational complexity increases factorially
Solution Approach 1:
The patent segments the solution space into multiple independent partitions, each assigned to a separate processing unit. Instead of one processor handling the entire solution space sequentially, the search space is divided such that multiple processors can simultaneously explore different regions, reducing the computational complexity from factorial to a more manageable level while maintaining complete coverage through systematic partitioning.
Solution Approach 2:
The patent introduces a parallel processing dimension to the traditional sequential search approach. By transitioning from a single-threaded sequential exploration to a multi-threaded parallel architecture, the system adds a temporal/spatial dimension to the search process, allowing multiple solution paths to be explored concurrently rather than one after another, thus reducing overall computational complexity.
2Speed
If sequential heuristics-based techniques are used to reduce computational complexity, then processing speed is improved, but cache utilization and computational resource usage increase significantly
Solution Approach 1:
The patent divides the solution space into discrete partitions that can be independently processed, allowing each processing unit to work on a specific segment with its own local cache. This segmentation reduces the need for large centralized cache memory, as each processor only needs to store and manage data for its assigned partition rather than the entire solution space.
Solution Approach 2:
The patent creates multiple copies of the search algorithm running in parallel on different processing units, each handling a specific partition of the solution space. Rather than one sequential algorithm requiring large cache to store all intermediate results, multiple simplified algorithm copies work simultaneously with smaller local caches, reducing overall memory requirements while maintaining processing speed.
3Productivity
If parallel processing is used to reduce computational resource usage and increase speed, then productivity is improved, but device complexity increases due to coordination overhead
Solution Approach 1:
The patent employs segmentation to create independent, non-overlapping partitions of the solution space, which eliminates the need for complex synchronization mechanisms between processing units. Each processor works autonomously on its assigned partition without requiring frequent communication or coordination, thereby reducing the overhead complexity associated with parallel processing while maintaining high productivity.
4Adaptability or versatility
If genetic and evolutionary searching is used to explore diverse solutions, then solution diversity is improved, but computational resource requirements increase
Solution Approach 1:
The patent distributes genetic and evolutionary searching operations across multiple processing units, each handling a specific partition of the solution space. This segmentation allows diverse solution exploration to occur in parallel rather than sequentially on a single processor, reducing the computational resource burden on any single unit while maintaining overall solution diversity through the combined results from all processors.
Data Source
AI summary
Systems and methods in accordance with the present disclosure can implement a parallel processing system, such as a graphics processing unit (GPU)-based system, to generate solutions to complex computational problems. Aspects of this technical solution can retrieve a plurality of solutions each representing a plurality of values in a multi-dimensional space, allocate, to one or more processing units associated with the one or more circuits and having a parallelized configuration, one or more of the plurality of solutions, modify, by the one or more processing units according to the parallelized configuration, at least one value of the one or more solutions allocated to the one or more processing units, to determine a plurality of modified solutions, and output, from the plurality of modified solutions, according to one or more criteria indicating a diversity among the plurality of modified solutions, a selected solution.