Lam Research had six patents in artificial intelligence during Q1 2024. Lam Research Corp has filed patents for methods and systems using machine learning to identify endpoints in semiconductor fabrication processes, accelerate recipe definition, and adaptively train models. Additionally, a robot system for servicing semiconductor tools has been developed, featuring a cart frame, arm support frame, arm frame, and arm locking mechanism for efficient operation. GlobalData’s report on Lam Research gives a 360-degree view of the company including its patenting strategy. Buy the report here.
Lam Research grant share with artificial intelligence as a theme is 16% in Q1 2024. Grant share is based on the ratio of number of grants to total number of patents.
Recent Patents
Application: Machine-learning in multi-step semiconductor fabrication processes (Patent ID: US20240096713A1)
The patent filed by Lam Research Corp. describes methods and systems for utilizing a time-series of spectra to determine the endpoint of multi-step semiconductor fabrication processes like deposition and etching. The approach involves creating a virtual carpet, which is essentially a machine learning model formed from a time-series of spectra collected during training operations. During production, in-situ spectra are compared to this virtual carpet to pinpoint the endpoint of multi-step fabrication processes. The patent outlines a method for generating a machine learning model that predicts substrate parameter values based on spectral data collected during multi-step processes, enabling real-time monitoring and control of these processes.
The patent's claims detail the steps involved in generating the machine learning model, including receiving training data with spectral data and parameter values, extracting features from the spectral data, and using this information to predict substrate parameter values for test substrates. The method is applicable to various multi-step processes, such as atomic layer etching or plasma etching, and can involve non-contiguous or contiguous steps. Additionally, the patent covers adjusting process conditions based on the machine learning model's output, demonstrating the potential for improved control and optimization of multi-step fabrication processes. Overall, the patent presents a comprehensive approach to leveraging machine learning and spectral data for enhanced monitoring and control of semiconductor fabrication processes.
To know more about GlobalData’s detailed insights on Lam Research, buy the report here.
Data Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.