Epitaxial Growth

IVW made

Artificial Intelligence Platform DOMM ™

IVWorks presents DOMM™, a deep learning-based artificial intelligence epitaxy platform that combines the latest
artificial-intelligence–based image analysis technology with hybrid-MBE technology. IVWorks’ outstanding semiconductor material growth experts and
artificial intelligence experts with domain knowledge collaborate to understand data better and create more robust artificial intelligence
prediction models using their ideas.

Machine learning platform for Epitxaial Wafer Production

The crystal structure information at the atomic layer level acquired in real time is analyzed and applied using deep-learning–based artificial-intelligence technology by the RHEED monitoring technique of hybrid-MBE to maximize yield rates.

AI collects, manages, and learns diffraction patterns reflected after focusing an electron beam on the thin film surface and the sensed data obtained from various equipment during the thin film synthesis process.

DOMM™ observes changes in sub-angstrom units and observes various pattern changes to learn and predict image information to apply information to the production process.

Architecture

Machine learning prediction model required for compound semiconductor growth

01
RSM

Sheet Resistance
Model
02
PDM

Pattern Detection
Model
03
CPM

Crystal Prediction
Model
04
ADM

Anomaly Detection
Model
05
RRM

Recipe Recommend
ation Model

DOMM™ was created to improve the productivity and competitiveness of compound semiconductor growth.

DOMM™ is a machine learning prediction model required for epitaxial growth to realize an artificial intelligence molecule-synthesis control system through the condition control, quality prediction, and real-time monitoring of the production process.

Level of AI Epitaxy System

	Level 0	Level 1	Level 2	Level 3	Level 4	Level 5
	Manual	Automatic	Partial AI Assistance	AI Assistance	AI Production	AI Manufacturing
Reclpe Development
Operation
Monitoring
Pass/Fail Prediction
Reclpe Correction