Aged Timing Model Generation for Circuit Aging Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for analyzing circuit designs to account for aging effects are either computationally costly or inaccurate, with existing techniques either incurring high overhead during cell characterization or resulting in pessimistic predictions of circuit performance.
Innovation Solution
A method that uses an analytical approach to generate an aged timing model by determining aging parameters for each cell type in a circuit design, allowing for the prediction of aging effects under target stress conditions without the need for costly characterization at those conditions, thereby reducing processing resources and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stress conditions are applied during cell characterization to generate an aged timing model, then the accuracy of aging prediction is improved, but the computational overhead during cell characterization increases significantly
Solution Approach 1:
The patent applies preliminary action by performing cell characterization with stress conditions only once during library generation, and then reusing the pre-computed aged timing models for multiple circuit designs. This eliminates the need to repeatedly perform costly stress-based characterization for each new circuit, thus improving accuracy while reducing computational overhead in subsequent analyses.
Solution Approach 2:
The patent creates copies of aged timing models from a single stress-based characterization and reuses these copied models across multiple circuit designs. Instead of performing expensive stress characterization on each new circuit, the system copies and applies the pre-generated aged timing models, significantly reducing computational overhead while maintaining prediction accuracy.
2Loss of energy
If an aging guardband is added to a fresh timing model during timing analysis, then the computational cost is reduced, but the accuracy and reliability of the timing prediction deteriorates due to large pessimism
Solution Approach 1:
The patent changes the parameters of the timing model by generating separate aged timing models with different stress conditions (temperature, voltage, time) instead of using a single fresh timing model with a guardband. This allows the system to capture non-linear aging effects across varying operating conditions, improving prediction accuracy without incurring the high computational cost of stress-based characterization for every circuit analysis.
Solution Approach 2:
The patent performs preliminary generation of aged timing models under various stress conditions during library characterization, and then reuses these pre-computed models during circuit timing analysis. This preliminary action eliminates the need to add conservative guardbands during each timing analysis, providing accurate aging predictions at low computational cost.
3Reliability
If stress conditions are applied during cell characterization, then the reliability of aging effects prediction is improved, but the time required for cell characterization increases
Solution Approach 1:
The patent performs stress-based cell characterization as a preliminary action during library generation, creating aged timing models that are then reused for multiple circuit designs. This one-time preliminary characterization improves the reliability of aging predictions while avoiding repeated time-consuming stress simulations for each new circuit analysis.
Solution Approach 2:
The patent copies the aged timing models generated from stress-based characterization and applies them to multiple circuit designs without re-running the expensive stress characterization. This copying approach maintains reliable aging predictions while significantly reducing the time required for subsequent circuit analyses.
Data Source
AI summary
A system and method for performing circuit design analysis obtains a circuit design comprising cells. The cells are associated with cell types. Aging parameters of a core analytical model are determined for each of the cell types in the circuit design to generate a calibrated analytical model. Aging effects for the cells are determined based on the calibrated analytical model and target stress conditions. An aged timing model is determined for the cell types based on the aging effects, an unaged timing model, and the target stress conditions.