Machine Learning Adaptation in Post Silicon Server Validation
Pridhiviraj Paidipeddi, Dheerendra Singh Tomar · International Journal of Applied Information Systems · 2014
As the complexity of modern server processors increases so the validation challenges.Current design validation methods cover less, resulting in bug escape and more regress postsilicon validation.The biggest problem is manual debugging of several failures by large number of test cases.By using machine learning in server validation, validation efforts and resource requirement will reduce.Validation of future generation server will be done through the learning set generated from the previous generation device, which is a set of test cases being passed.