Description
The role develops a formal-methods-based framework for generating legal and realistic stimulus for custom machine-learning accelerator chips. Responsibilities include understanding the accelerator instruction set, modeling instruction-stream correctness with first-order logic, creating APIs for scheduling and randomness, deploying algorithms for safe concurrent code, establishing coverage metrics, and using machine learning to automate content creation. The position requires at least three years of business-application modeling experience, a PhD or master's degree with four or more years of relevant experience, patent or publication experience, and programming experience in Java, C++, Python, or a related language. The role is based in Cupertino, California, with a base salary of 171,600–222,200 USD annually.
