Power Modeled Integrated Circuits Using Slew-Sensitive Transfer Functions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for modeling cross current power consumption in integrated circuits are inadequate, as logic simulation fails to account for input slew dependence, while circuit simulation is computationally intensive.
Innovation Solution
A method involving logic simulation, sensitivity transfer functions, and circuit simulation is employed to model cross current power consumption, using representative cells with common family classifications to generate sensitivity transfer functions, reducing the need for extensive circuit simulations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If logic simulation is used to model power consumption, then computational resources are reduced, but accuracy is insufficient because it fails to account for input slew dependence
Solution Approach 1:
A sensitivity transfer function is introduced as an intermediary mathematical model that bridges logic simulation and circuit simulation. This transfer function captures the input slew dependence relationship derived from circuit simulation results, allowing logic simulation to achieve both low computational cost and high accuracy by incorporating the slew sensitivity characteristics without performing expensive circuit simulations for every cell.
2Measurement precision
If circuit simulation is performed on all cells, then power consumption modeling accuracy is improved, but computational intensity increases significantly
Solution Approach 1:
The methodology segments the IC design into representative cells that capture the essential power consumption characteristics. Instead of performing circuit simulation on all cells, only a subset of representative cells is simulated to generate sensitivity transfer functions. These transfer functions are then applied to model power consumption for all cells in the design, significantly reducing total computational intensity while maintaining accuracy.
Solution Approach 2:
The approach changes the modeling parameters from direct circuit simulation results to sensitivity transfer functions that characterize input slew dependence. By transforming the power consumption model to use sensitivity parameters derived from representative cells, the system achieves accurate power modeling across all cells without the computational burden of full circuit simulation for each cell.
3Loss of energy
If representative cells with common family classification are used, then computational resources are reduced, but the complexity of classification and transfer function generation increases
Solution Approach 1:
The sensitivity transfer function serves multiple functions: it captures power consumption characteristics, accounts for input slew dependence, and enables prediction across different cell instances within a family. By creating a universal transfer function model that can be applied to all cells of a given family type, the system reduces the need for individual cell analysis while maintaining accuracy across the entire cell library.
Data Source
AI summary
Methods, computer program products, and systems are presented. The method computer program products, and systems can include, for instance: performing logic simulation of a cell and determining, in dependence on the performing logic simulation of the cell, a first at least one component of power consumption of the cell, wherein the cell is included in a transistor level IC design selected for modeling; determining, with use of a sensitivity transfer function, a second at least one component of power consumption of the cell, wherein generating of the sensitivity transfer function has included subjecting a representative cell to circuit simulation, the representative cell having a common family classification with the cell; and ascertaining a power consumption of the cell in dependence on the first at least one component of power consumption of the cell and in dependence on the second at least one component of power consumption of the cell.