Statistical Timing Analysis Parameter Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional statistical static timing analysis (SSTA) methods require multiple timing runs and introduce non-linearity errors when modeling parameter variations, leading to reduced process subspace coverage and increased complexity in IC chip design and fabrication.
Innovation Solution
The method involves splitting parameters of interest into controlled and uncontrolled parts, correlating or anti-correlating the controlled part with another parameter, and projecting timing results using these correlations to achieve improved modeling of variability and coverage of a larger process subspace with fewer timing runs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If first-order linear approximation (canonical model) is used for SSTA, then computational simplicity is maintained, but non-linearity errors are introduced when parameter range increases
Solution Approach 1:
The parameter of interest is divided into two separate parameters: a controlled parameter and an uncontrolled parameter. This segmentation allows the controlled parameter to be used for timing closure with reduced non-linearity errors, while the uncontrolled parameter accounts for residual variations, thereby maintaining both computational simplicity and timing analysis precision.
2Measurement precision
If multiple timing runs are performed to cover full parameter range, then timing closure accuracy is improved, but computational time and complexity increase
Solution Approach 1:
By splitting the parameter of interest into controlled and uncontrolled parts, the invention enables a single timing run to cover the full parameter range effectively. The controlled parameter is used for precise timing closure while the uncontrolled parameter captures remaining variations, eliminating the need for multiple timing runs and reducing computational time.
Solution Approach 2:
The invention transforms a single parameter into two parameters with different characteristics (controlled and uncontrolled). This parameter transformation allows the timing analysis to efficiently cover the entire parameter space in one run by leveraging the controlled parameter for accurate timing closure and the uncontrolled parameter for comprehensive coverage.
3Loss of time
If nth-order extension with cross terms is used for timing closure, then single timing run is sufficient, but process subspace coverage is significantly reduced
Solution Approach 1:
The invention segments the parameter of interest into controlled and uncontrolled components, where the controlled parameter enables single-run timing closure (reducing computational time) while the uncontrolled parameter maintains extensive process subspace coverage (improving adaptability). This segmentation resolves the contradiction between efficiency and coverage.
Data Source
AI summary
Systems and methods for improving timing closure of new and existing IC chips by breaking at least one parameter of interest into two or more partial parameters. More specifically, a method is provided for that includes propagating at least one timing analysis run for a semiconductor product. The method further includes identifying at least one parameter of interest used in the at least one timing analysis run. The method further includes splitting the at least one parameter into two parts comprising a controlled part and an uncontrolled part. The method further includes correlating or anti-correlating the controlled part with another parameter used in the at least one timing analysis run. The method further includes projecting timing using the correlation or anti-correlation between the controlled part and the another parameter and using the uncontrolled part of the at least one parameter.