About MST Semiconductor
We build AI that lives on the fab floor — accelerating semiconductor equipment from design through volume production.
Our Mission
Semiconductor equipment development is one of the most complex engineering challenges in the world. Yet much of the workflow — from mechanical design to process tuning to production control — still depends on manual iteration and deep tribal knowledge.
MST Semiconductor (Shanghai) Co., Ltd. was founded to change that. Our principle is simple: AI-Defined Manufacturing. We bring machine learning and intelligent automation to every stage of the equipment lifecycle, so engineering teams can move faster, waste less, and scale with confidence.
What We Do
We develop NeuroBox, a suite of AI-native solutions purpose-built for semiconductor equipment companies and fabs. NeuroBox covers three critical stages:
Design — NeuroBox D
AI-powered design automation that learns from your existing assemblies and standard part libraries. Generate production-ready 3D CAD models from P&ID inputs, cutting mechanical design cycles by an order of magnitude.
Commissioning — NeuroBox E5200
Smart DOE for equipment bring-up and process qualification. Our models identify optimal operating windows with up to 80% fewer test wafers, turning weeks of trial-and-error into days of systematic convergence.
Production — NeuroBox E3200
Real-time production AI for virtual metrology, run-to-run control, and equipment intelligence. Sub-50ms inference on live process data keeps your tools running at peak performance, wafer after wafer.
On-Premise by Design
Semiconductor manufacturing data is among the most sensitive IP in any industry. That is why every NeuroBox product runs entirely on-premise, on edge computing infrastructure deployed inside your facility. Your process data, equipment parameters, and model outputs never leave the fab. There is no cloud dependency, no external data transfer — just AI that works where your tools work.
Our Team
MST Semiconductor is led by a PhD-level team at the intersection of semiconductor process engineering and applied AI. Our researchers and engineers bring hands-on experience from equipment development, fab operations, and production-scale machine learning — the combination required to build AI that actually works in a cleanroom environment.
We operate dual R&D centers in Shanghai: our core AI research lab in Zhangjiang, the heart of China’s semiconductor corridor, and our systems engineering center in Jiading, focused on hardware integration and deployment.
Who We Work With
Our technology is deployed with leading semiconductor equipment manufacturers and production fabs across China. We partner closely with equipment OEMs and fab engineers to ensure NeuroBox delivers measurable results in real production environments — not just in the lab.
Work With Us
Whether you are building the next generation of semiconductor equipment or running a high-volume production line, we would like to show you what AI-Defined Manufacturing looks like in practice.
