MACHINE LEARNING BASED SMART PROCESS RECIPE BUILDER TO IMPROVE AZIMUTHAL FLOW AND THICKNESS UNIFORMITY

MACHINE LEARNING BASED SMART PROCESS RECIPE BUILDER TO IMPROVE AZIMUTHAL FLOW AND THICKNESS UNIFORMITY

Methods, software systems and processes to develop surrogate model-based optimizers for controlling and optimizing flow and pressure of purges between a showerhead and a heater having a substrate support to control non-uniformity inherent in a processing chamber due to geometric configuration and pr...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Shah, Kartik, AuBuchon, Joseph, Griffin, Kevin, Ramanathan, Karthik, Baluja, Sanjeev, Chen, Hanhong, Wang, Chaowei, Kashyap, Dhritiman Subha
Format: Patent
Sprache:eng
Schlagworte:
CALCULATING
COMPUTING
COUNTING
ELECTRIC DIGITAL DATA PROCESSING
PHYSICS
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Methods, software systems and processes to develop surrogate model-based optimizers for controlling and optimizing flow and pressure of purges between a showerhead and a heater having a substrate support to control non-uniformity inherent in a processing chamber due to geometric configuration and process regimes. The flow optimizer process utilizes experimental data from optimal process space coverage models, generated simulation data and statistical machine learning tools (i.e. regression models and global optimizers) to predict optimal flow rates for any user-specified process regime.