Methods and mechanisms for modifying machine-learning models for new semiconductor processing equipment
AssigneeApplied Materials, Inc.
InventorsWaheb Bishara, Subhasish Bandyopadhyay
Life of this patent
Estimating the term…
Abstract
An electronic device manufacturing system configured to identify, by a processor, a first machine-learning model trained to generate predictive data for a first process chamber and obtain metrology data associated with a substrate produced by a second process chamber. The manufacturing system is further configured to train a second machine-learning model based on the first machine-learning model and the metrology data, wherein the second machine-learning model is trained to generate predictive data associated with the second process chamber.
Claims20
A method, comprising:
- identifying a first machine-learning model trained to generate first predictive data for a first process chamber;
- obtaining metrology data associated with a substrate processed by a second process chamber;
- and
- training a second machine-learning model based on the first machine-learning model and the metrology data by performing a transfer learning operation, wherein data associated with the first machine-learning model is used to train the second machine-learning model, and, wherein the second machine-learning model is trained to generate second predictive data associated with the second process chamber.
An electronic device manufacturing system, comprising:
- a memory device;
- and
- a processing device, operatively coupled to the memory device, to perform operations comprising:
- identifying a first machine-learning model trained to first generate predictive data for a first process chamber;
- obtaining metrology data associated with a substrate processed by a second process chamber;
- and
- training a second machine-learning model based on the first machine-learning model and the metrology data by performing a transfer learning operation, wherein data associated with the first machine-learning model is used to train the second machine-learning model, and, wherein the second machine-learning model is trained to generate second predictive data associated with the second process chamber.
A method, comprising:
- providing input data to a first machine-learning model trained to generate first predictive data for a first process chamber, the input data comprising a value associated with a knob that controls a substrate manufacturing process, wherein the first machine-learning model is trained by performing a transfer learning operation in which data associated with a second machine-learning model trained to generate second predictive data for a second process chamber is used to train the first machine-learning model, and metrology data associated with a prior substrate processed by the first process chamber;
- and
- obtaining one or more output values of the first machine-learning model, the one or more output values reflecting a predictive dimension on a substate.
Dependencies are inferred from the claim text.