Memory Device Virtual Quality Control Using Area-Based Test Prediction

The patent applies preliminary action by performing virtual testing and interpolation of quality control data before actual physical testing. The system uses machine learning models to predict test results based on process parameters and previously collected data, allowing manufacturers to pre-determine which samples need physical testing and what thresholds to apply, thereby reducing the time required for extensive physical testing while maintaining accuracy.

The patent implements copying by creating virtual copies of test results through interpolation and machine learning predictions. Instead of physically testing every sample, the system generates virtual quality control data that replicates the information obtained from actual testing. This virtual data copying approach maintains measurement precision while significantly reducing the time and resources required for physical testing.

The system performs preliminary virtual testing and data interpolation before physical testing, allowing manufacturers to quickly assess quality control metrics without waiting for time-consuming physical tests. This preliminary action enables faster decision-making and maintains manufacturing throughput while ensuring measurement precision through subsequent targeted physical testing of only necessary samples.

By creating virtual copies of quality control data through machine learning and interpolation, the system enables rapid quality assessment that does not bottleneck manufacturing productivity. The virtual data copying allows parallel processing and faster feedback loops, maintaining both measurement precision and high manufacturing throughput.

The patent applies copying by generating virtual test data that replicates the information from physical testing without consuming physical samples. The interpolation algorithms create virtual copies of quality control measurements for samples that were not physically tested, thereby increasing the volume of available test data while preserving the usability of physical samples for other purposes or for confirmatory testing.

The system replaces the mechanical physical testing process with a computational virtual testing system. Instead of physically testing every sample to generate extensive data, the system uses machine learning models and interpolation algorithms to generate virtual test data, substituting computational processes for physical testing and thereby preserving sample usability while maintaining data volume.

The system performs preliminary virtual quality control assessment using machine learning models trained on historical data and process parameters. This preliminary action provides reliable quality control insights quickly, allowing manufacturers to make rapid decisions about which samples require detailed physical testing, thereby maintaining reliability while reducing overall testing cycle time.

By creating virtual copies of quality control data through interpolation, the system enables rapid reliability assessment without time-consuming physical testing of every sample. The virtual data copying maintains reliability by using robust statistical methods and machine learning models that accurately predict quality outcomes, significantly reducing testing cycle time while preserving reliability.