Onur Atan

Title: Onur Atan and Machine Learning for Netlist Design.

Introduction

Onur Atan is an inventor based in San Diego, California, and is associated with Qualcomm Incorporated. His work includes a patent application describing the use of machine learning to evaluate electronic circuit designs.

Latest Patents

Atan is credited with 1 patent. His patent, “Machine Learning for Netlist Design,” describes techniques for assessing an electrical circuit’s netlist design. The approach represents the design as a directed graph, selects a node corresponding to a register, and generates a subgraph that captures the register’s connectivity cone. A trained machine-learning encoder then creates a functional embedding of that subgraph, which is used to predict a performance characteristic of the netlist design.

Career Highlights

Atan’s disclosed invention applies machine learning to electronic design evaluation. By deriving a functional representation from circuit connectivity, the technique aims to support prediction of design performance.

Collaborations

Atan is associated with Qualcomm Incorporated. Gokce Sarar is identified as a coworker.

Conclusion

Onur Atan’s patent work brings together circuit-design analysis and machine learning to predict performance characteristics from netlist structures.