Circuit Clock Skew Optimization via Current Source Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Circuits face challenges in minimizing variance of waveform arrival times due to process voltage temperature (PVT) variations, leading to clock skew issues, which are exacerbated in deep submicron technology with high clock frequencies.
Innovation Solution
A method and system that define an objective function using the shape of waveforms measured at timing points in each branch of a circuit, optimizing it to minimize clock skew variance across different PVT values by generating a current source model, quantifying waveforms as linear combinations of independent basic shapes, and optimizing coefficients to reduce skew while maintaining similar waveform shapes across branches, involving buffer insertion and gate sizing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional circuit design is used without optimization, then device complexity is low, but clock skew variance increases under PVT variations
Solution Approach 1:
The patent performs preliminary waveform shape analysis and current source model generation before final circuit optimization. By pre-characterizing the waveform shapes at different PVT corners and creating current source models, the optimization process can focus on specific parameters rather than exploring the entire design space, thus reducing complexity while improving clock skew variance.
Solution Approach 2:
The patent replaces traditional trial-and-error or exhaustive optimization methods with a physics-based current source model approach. By modeling the waveform shapes using current source characteristics and using linear combination of basis waveforms, the system substitutes complex iterative simulation with a more efficient mathematical framework that reduces computational complexity.
2Reliability
If circuit optimization is performed to minimize clock skew variance, then reliability improves, but manufacturing precision requirements increase
Solution Approach 1:
The patent transforms the optimization problem from matching absolute waveform parameters to matching waveform shape characteristics. By using current source models and basis waveform linear combinations, the optimization focuses on shape parameters (temporal distribution patterns) rather than absolute timing values, which are more sensitive to manufacturing variations. This approach reduces manufacturing precision requirements while maintaining reliability.
3Reliability
If waveform shapes are matched across branches, then clock skew variance decreases, but measurement and detection difficulty increases
Solution Approach 1:
The patent creates simplified copies of complex waveforms using current source models and basis waveform linear combinations. Instead of directly measuring and comparing full waveform shapes (which is difficult), the system creates mathematical representations (copies) that capture the essential shape characteristics. These models can be easily manipulated and compared, reducing measurement difficulty while maintaining the benefit of waveform shape matching.
Data Source
AI summary
A method for optimizing a circuit includes at least a first branch and a second branch includes defining an objective function using a shape of waveforms measured at a timing point in each branch, and optimizing the objective function to minimize a variance of clock skew of the first branch and the second branch across different process voltage temperature values.