Parallel Solver Architecture for Faster Combinatorial Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems face challenges in efficiently solving combinatorial optimization problems due to factorial complexity, limited parallelization, and resource constraints, leading to reduced accuracy and optimal solutions.
Innovation Solution
The implementation of parallel processing techniques using multiple parallel processing units (PPUs) to execute various tasks, such as globalizing heuristics and hill climbers, to determine high-quality solutions for combinatorial optimization problems, leveraging algorithms, metaheuristics, and artificial intelligence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sequential processing approaches are used to solve combinatorial optimization problems, then computational resources and time are conserved, but the search space explored is limited to a small fraction and solution accuracy deteriorates
Solution Approach 1:
The patent divides the combinatorial optimization problem into multiple independent sub-problems or search tasks that can be processed in parallel. Each processing unit explores a different portion of the search space simultaneously, enabling comprehensive coverage without requiring sequential evaluation of all possibilities. This segmentation allows the system to maintain high solution accuracy while reducing overall computation time through concurrent exploration of multiple solution paths.
2Measurement precision
If the search space is expanded to explore more potential solutions, then solution accuracy improves, but computational intensity and time increase excessively
Solution Approach 1:
The patent segments the large search space into multiple smaller, manageable portions that can be explored in parallel by different processing units. This segmentation enables the system to expand the effective search space coverage without proportionally increasing computational intensity for each individual unit, as the workload is distributed across multiple concurrent processors rather than concentrated in a single sequential processor.
Solution Approach 2:
The patent combines the results from multiple parallel processing units to generate the final optimized solution. By merging the findings from numerous concurrent explorations of the search space, the system achieves comprehensive coverage and high solution accuracy while maintaining efficient resource utilization through parallel rather than sequential processing.
3Productivity
If conventional processing systems are used, then device complexity is low, but the ability to parallelize algorithms is limited and productivity deteriorates
Solution Approach 1:
The patent segments the computation into independent parallel tasks that can be executed simultaneously by multiple processing units. This segmentation enables high productivity through parallel execution while managing device complexity by organizing the parallel architecture around standardized, reusable processing modules that handle specific aspects of the optimization problem.
Solution Approach 2:
The patent employs universal processing units that can handle multiple aspects of the combinatorial optimization problem through parallel execution. These multi-functional processing units reduce overall system complexity by using identical or similar hardware components for different computational tasks, rather than requiring specialized hardware for each function, thereby improving productivity without proportionally increasing device complexity.
Data Source
AI summary
In various examples, solutions to combinatorial optimization problems are determined using a plurality of solvers executing in parallel. In an embodiment, the plurality of solvers executed in parallel perform one or more search algorithms. Furthermore, in such embodiments, the operations of the one or more search algorithms are also executed in parallel.