Corner Database Generator for Memory Compiler Customization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional memory compilers lack the capability and flexibility for users to generate new corner databases independently, requiring vendors to create and ship new corners, which limits user autonomy and efficiency in memory instance model generation.
Innovation Solution
A system and method for generating new corner databases using existing memory compiler corner databases and minimum characterized data, enabling users to define and create new corners through a computing device with a memory compiler corner database generator, which includes a mapper and a corner database builder to produce memory instance structures and EDA views for the new corner.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional memory compiler is used, then corner databases are generated by vendor, but user flexibility and autonomy are limited
Solution Approach 1:
The system enables users to generate their own corner databases using the corner database generator, eliminating the need to rely on vendor-provided databases. Users can independently create customized corner databases by providing their own process corner data, thereby achieving self-service and improved flexibility.
Solution Approach 2:
The corner database generator acts as an intermediary tool between raw process corner data and the final corner database. It processes and transforms user-provided data into usable corner databases, bridging the gap between simple data input and complex database generation while maintaining system manageability.
2Productivity
If vendor generates and ships new corners, then corner database quality is ensured, but user autonomy and efficiency are reduced
Solution Approach 1:
The system allows users to customize corner database parameters such as process corner characteristics, voltage levels, and temperature conditions according to their specific needs. By enabling parameter customization, users can efficiently generate targeted corner databases while maintaining data quality through controlled parameter specification.
Solution Approach 2:
Users prepare and provide their own process corner data and specifications in advance before generating the corner database. This preliminary preparation ensures that the input data meets quality requirements, enabling efficient generation while maintaining reliability through pre-validated inputs.
3Adaptability or versatility
If custom corner is needed, then user-specific requirements are met, but additional vendor work and shipping time are required
Solution Approach 1:
Users can immediately generate customized corner databases using the corner database generator without waiting for vendor processing. The self-service capability allows users to create custom corners on-demand, eliminating vendor turnaround time and enabling immediate adaptation to specific design requirements.
Solution Approach 2:
The corner database generator serves as a local intermediary tool that users can execute on their own systems. This eliminates the need for external vendor intervention and shipping, allowing rapid customization of corner databases directly at the user site without time loss.
Data Source
AI summary
Various implementations described herein are directed to a computing device. The computing device may include a mapper module that receives a user configuration input of a destination corner for building a destination corner database. The mapper module may include a decision making engine that decides fabrication parameters for building the destination corner database based on the verified user configuration input and memory compiler metadata. The computing device may include a builder module that performs a simulation of the destination corner based on the fabrication parameters, collects simulation results data associated with the simulation, and builds the destination corner database for the destination corner based on the simulation results data and source corner data. The computing device may include a memory compiler that accesses the destination corner database and generates memory instance structures and their electronic digital automation (EDA) views for the destination corner based on the destination corner database.