IC Timing Estimation Using Analytic Cell Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining the post-physical-optimization timing of IC designs require physical optimization, which is time-consuming and costly, and do not provide early feedback on design feasibility, leading to delayed design-cycle surprises and inefficiencies in redesign and iteration.
Innovation Solution
A system that estimates post-physical-optimization timing using analytic models of cells and netlists, allowing for the prediction of timing without performing physical optimization, by generating load-delay and load-capacitance models and simulating buffer chain and repeater insertion to calculate delays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physical optimization is performed to determine post-physical-optimization timing, then timing accuracy is improved, but turnaround time increases significantly
Solution Approach 1:
The patent creates a simplified analytical model that copies the essential characteristics of physical optimization without performing the complete physical optimization process. The analytical timing model replicates the timing prediction functionality of physical optimization while using simplified calculations based on pre-computed cell models and netlist information, achieving acceptable accuracy without the computational burden of full physical optimization.
Solution Approach 2:
The patent performs preliminary actions by pre-computing analytical models for each cell type during library characterization. These pre-computed models capture the timing behavior of cells under various conditions. During actual timing analysis, these pre-computed models are applied directly to the netlist, eliminating the need to perform complex physical optimization calculations at runtime, thus significantly reducing turnaround time while maintaining accuracy.
2Reliability
If physical optimization is performed to ensure timing closure, then design reliability is improved, but the number of design iterations increases
Solution Approach 1:
The patent implements a feedback mechanism where the analytical timing model provides rapid timing predictions that can be used to evaluate design decisions before committing to full physical optimization. This early feedback allows designers to identify potential timing issues and adjust their designs accordingly, reducing the likelihood of failing timing closure after expensive physical optimization iterations, thus improving overall design reliability and reducing iteration counts.
Solution Approach 2:
By performing preliminary timing analysis using the analytical model, designers can identify and fix timing violations early in the design process, before committing to full physical optimization. This preliminary action prevents wasted iterations on designs that would fail timing closure anyway, improving both reliability and productivity.
3Productivity
If early timing prediction is implemented without physical optimization, then productivity is improved, but measurement precision deteriorates
Solution Approach 1:
The analytical timing model creates a simplified copy of the physical optimization timing prediction process. By pre-computing cell models that capture essential timing characteristics and using these models to predict timing from netlist information alone, the system achieves productivity improvement while maintaining sufficient accuracy for design exploration and budgeting purposes.
Solution Approach 2:
The patent changes the parameters used for timing prediction from detailed physical optimization results to simplified analytical model parameters derived from netlist information and pre-computed cell characteristics. This parameter transformation enables fast prediction while maintaining acceptable accuracy for early-stage design decisions.
Data Source
AI summary
A system that determines the timing of an integrated circuit (IC) design is presented. During operation, the system receives a netlist for the IC design, wherein the netlist specifies the placement of cells within the IC design. Next, the system estimates capacitances for cells within the IC design based on analytic models of the cells. The system then estimates the post-physical-optimization timing of the IC design based on the netlist, the capacitances, and the analytic models, wherein the post-physical-optimization timing is estimated without performing physical optimization.