Cutting-edge insights and technical practices in semiconductor intelligent manufacturing
Key TakeawayAI-enhanced OES (Optical Emission Spectroscopy) monitors plasma state across hundreds of wavelength channels in real-time. Machine learning identifies etch endpoints and process anomalies with higher accuracy than traditional threshold methods. Plasma OES Monitoring: Etch Endpoint Detection and Process Anomaly...
Read more →
Key TakeawayDomestic semiconductor equipment manufacturers can accelerate AI adoption through third-party platforms like NeuroBox. While international leaders (KLA, ASML, Applied Materials) have built-in AI, domestic OEMs can achieve equivalent capabilities in 2-4 weeks. Semiconductor Equipment Localization: How AI Accelerates the...
Read more →
Key TakeawayNeuroBox E5200 Technical Brief: AI-powered Smart DOE system reducing trial wafers by 80%. Built on NVIDIA Jetson Orin NX with Bayesian optimization, Latin hypercube sampling, and transfer learning. Deploys in 2-4 weeks for equipment commissioning. MST Semiconductor · Technical...
Read more →
Key TakeawayAI improves lithography overlay control to sub-2nm accuracy required for 7nm and below. Multi-dimensional data fusion (equipment parameters, environmental data, historical trends) enables predictive compensation instead of reactive feedback. Lithography Overlay Control: How AI Enhances Alignment Accuracy Author: MST...
Read more →
Key TakeawayNeuroBox E3200 Technical Brief: Edge AI computing platform for semiconductor production lines. Provides VM, R2R, and EIP with sub-50ms latency. Based on NVIDIA Jetson Orin, supporting SECS/GEM and 300+ equipment models. MST Semiconductor · Technical Whitepaper NeuroBox E3200: Edge...
Read more →
Key TakeawayNeuroBox D Technical Brief: AI-powered mechanical design platform generating SolidWorks assemblies from P&ID. Learns from customer CAD libraries, reduces design time by 60%, and produces native .sldasm files with BOM and routing. MST Semiconductor · Technical Whitepaper NeuroBox D:...
Read more →
