- Improve test time by identifying highly correlated combinations of tests, from single to multiple test insertions, and removing redundant tests
- Loosen or tighten test limits in a robust manner to reduce test escapes or increase yield
- Identify outliers using bivariate and multivariate algorithms and automatically re-binning them to improved product quality
How It Works
EXACTLY what you need:
Characterization & HVM under one hood
Characterization and HVM may represent two distinct phases in the product lifecycle, but they are ultimately linked together on the quality spectrum; the more issues you can solve prior to production ramp-up, the less likely you are to run into preventable escapes and RMAs down the line.
What is the relationship between these two seemingly disparate phases? Both HVM and characterization are actually best performed through the auspices of a Big Data platform. In high volume manufacturing environments, engineers typically need to analyze millions of parts across multiple test insertions, which can generate billions of data points. In contrast, the characterization phase focuses on looking at only a limited amount of device where there are hundreds of thousands of tests and test conditions that need to be considered in order to ensure that products are of the highest quality standards prior to ramp up and production.