Netlist Optimization via BDD Sweeping and Rewriting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for optimizing netlist representations in logic design and verification are limited in scope, often focusing on local optimizations and increasing design size, which hampers computational efficiency and resource utilization.
Innovation Solution
The use of Binary Decision Diagrams (BDDs) for global optimization, where predictive heuristics identify reducible regions, and BDD sweeping and rewriting algorithms reduce the netlist size by replacing logic with optimized structures, while avoiding unnecessary resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If local optimization techniques are used to rewrite logic expressions, then computational complexity is reduced, but the scope of optimization is limited and netlist size reduction is insufficient
Solution Approach 1:
The patent segments the netlist into regions that can be independently optimized using BDDs, allowing global optimization to be achieved through coordinated local operations. This enables the system to manage complex optimizations in manageable portions while maintaining overall netlist reduction.
Solution Approach 2:
The patent transitions from traditional local, level-by-level optimization to a global, region-based optimization approach using BDDs. This dimensional shift in the optimization strategy enables simultaneous consideration of multiple logic levels and regions, achieving both broad scope and significant size reduction.
2Productivity
If BDDs are constructed for all portions of logic to achieve global optimization, then netlist size is reduced, but computational resources and time are excessively consumed
Solution Approach 1:
The patent applies BDD construction selectively to specific regions of the netlist rather than uniformly across the entire design. By identifying and optimizing only those regions that will benefit most from BDD-based rewriting, the system achieves effective netlist reduction while minimizing unnecessary computational resource consumption.
Solution Approach 2:
The patent performs partial BDD construction on selected logic portions rather than exhaustive construction on all logic. This partial action approach achieves sufficient netlist optimization without the excessive time and resource costs of complete global BDD construction.
3Use of energy by moving object
If existing BDDs are leveraged for rewriting, then resource allocation is efficient, but the degree of optimization is constrained by the quality of existing BDDs
Solution Approach 1:
The patent performs preliminary BDD construction and analysis on selected regions before the main optimization pass. This preliminary action identifies high-value optimization opportunities and prepares BDD representations in advance, enabling more effective rewriting while maintaining efficient resource allocation during the actual optimization process.
Solution Approach 2:
The patent uses feedback from analyzing existing BDDs to guide selective construction of additional or improved BDDs. By evaluating the quality and utility of existing BDD representations, the system determines where additional BDD construction will yield the most benefit, balancing resource efficiency with optimization effectiveness.
Data Source
AI summary
Leveraging existing Binary Decision Diagrams (BDDs) to enhance circuit reductions in a system model representing a state machine as a netlist. The netlist is evaluated to determine the regions with the greatest potential reductions. BDD sweeping is performed to identify redundancies in the netlist. BDD rewriting implements the circuit reductions by replacing gates of the original netlist with more efficient equivalent logic.