Monte Carlo Circuit Simulation Optimization via Device Merging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Monte Carlo circuit simulation methods face performance and memory limitations due to brute force approaches, which disable optimizations, limiting the number of devices that can be simulated and resulting in degraded performance and increased memory requirements.
Innovation Solution
The method optimizes Monte Carlo simulations by combining devices sharing common parameters into a single optimized device, using machine learning techniques to generate random variations and apply them to these optimized devices, thereby reducing computational load and memory needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If brute force methods are used to uniquely and randomly vary all devices, then simulation accuracy is improved, but memory requirements and computational performance degrade significantly
Solution Approach 1:
The patent merges multiple devices that share common parameters into a single optimized device representation. Instead of uniquely varying each device individually (brute force approach), the system combines devices with identical parameter structures into one representative device, thereby reducing the total number of unique device variations that need to be simulated while maintaining statistical accuracy through proper random variation application to the optimized device.
Solution Approach 2:
The patent transforms the simulation approach by changing how device parameters are handled. Rather than assigning unique random variations to each device (which explodes memory and computational requirements), the system applies parameter transformations to optimized devices that represent multiple original devices. This parameter-based optimization maintains the statistical properties needed for accurate Monte Carlo simulation while dramatically reducing resource requirements.
2Measurement precision
If brute force methods are used to uniquely and randomly vary all devices, then simulation accuracy is improved, but computational performance degrades significantly
Solution Approach 1:
The patent merges multiple devices that share common parameters into a single optimized device representation. Instead of uniquely varying each device individually (brute force approach), the system combines devices with identical parameter structures into one representative device, thereby reducing the total number of unique device variations that need to be simulated while maintaining statistical accuracy through proper random variation application to the optimized device.
Solution Approach 2:
The patent transforms the simulation approach by changing how device parameters are handled. Rather than assigning unique random variations to each device (which explodes memory and computational requirements), the system applies parameter transformations to optimized devices that represent multiple original devices. This parameter-based optimization maintains the statistical properties needed for accurate Monte Carlo simulation while dramatically reducing resource requirements.
3Productivity
If devices sharing parameters are combined into optimized devices, then simulation performance and capacity are improved, but device complexity in the netlist is reduced
Solution Approach 1:
The patent merges multiple devices that share common parameters into a single optimized device representation. Instead of uniquely varying each device individually (brute force approach), the system combines devices with identical parameter structures into one representative device, thereby reducing the total number of unique device variations that need to be simulated while maintaining statistical accuracy through proper random variation application to the optimized device.
Data Source
AI summary
A computer-implemented method for optimizing a circuit simulator and computing surrogate models using circuit theory-guided machine learning for high performance Monte Carlo simulations. The method may include reading a netlist for performing a Monte Carlo simulation of an electronic circuit, and enabling simulator optimization such that two or more devices sharing one or more device parameters can be combined and simulated as one single optimized device. The method may also include constructing equations of netlist parametric expressions for one or more optimized devices, and computing optimized device mappings.