Hardware Simulator Cell Segmentation for Simulation Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current simulation technologies for hardware, such as integrated circuits, suffer from inefficiencies like unnecessary evaluations and memory cache misses due to the lack of differentiation between active and inactive cells, leading to wasted processing time and resources.
Innovation Solution
A hardware simulator sorts cells into active and inactive categories based on changes in outputs, inputs, or internal states relative to a potential clock signal, using sub-arrays to track and manage these changes, thereby reducing unnecessary evaluations and memory access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If hardware simulators evaluate all cells during simulation, then complete circuit validation is achieved, but processing time is excessively wasted due to unnecessary evaluations of inactive cells
Solution Approach 1:
The patent segments the cell population into active and inactive subsets using separate arrays. Active cells are tracked in one array while inactive cells are tracked in another, allowing the simulator to selectively evaluate only active cells during simulation events. This segmentation resolves the contradiction by maintaining complete validation coverage (all cells are tracked) while eliminating wasted processing time (only active cells are evaluated).
Solution Approach 2:
The patent implements dynamic cell status tracking where cells can transition between active and inactive states based on simulation events. The simulator continuously updates cell status and adjusts which cells require evaluation, making the evaluation process adaptive rather than static. This dynamic approach ensures complete validation while minimizing unnecessary evaluations at each simulation step.
2Reliability
If hardware simulators access and update information for all cells, then comprehensive monitoring is achieved, but memory cache misses increase due to accessing inactive cells
Solution Approach 1:
The patent uses segmented arrays to separate active and inactive cells in memory. By organizing cell data into distinct arrays based on activity status, the simulator reduces random memory access patterns and improves cache utilization. Only active cells are accessed during evaluation, significantly reducing memory cache misses while maintaining comprehensive monitoring capability.
Solution Approach 2:
The patent extracts inactive cells from the active cell set and places them in a separate inactive array. This extraction eliminates the need to access and update information for inactive cells during simulation events, directly reducing memory cache misses while preserving the ability to monitor all cells for status changes.
3Speed
If hardware simulators evaluate cells frequently, then real-time circuit behavior is captured, but processing resources are wasted on cells with unchanged outputs
Solution Approach 1:
The patent implements dynamic cell status tracking where cells transition between active and inactive states based on whether their outputs have changed. This dynamic classification allows the simulator to capture real-time circuit behavior by evaluating only cells that have actually changed state, thereby maintaining simulation accuracy while improving effective processing throughput by eliminating redundant evaluations.
Solution Approach 2:
The patent enables cells to effectively self-identify their evaluation needs through status flags and change detection mechanisms. Cells that have not changed their output state automatically become inactive and are excluded from further evaluation until they change, allowing the simulation system to automatically adapt its processing focus without external intervention and improving overall productivity.
Data Source
AI summary
An electrical circuit comprising a plurality of cells can be simulated to produce simulation results by sorting cells between active status cells and inactive status cells and reducing the processing of simulation results from inactive cells to thereby save simulation time.