Understand. Design. Integrate.
VDLab investigates phenomena that span scales, turns that understanding into virtual models for materials, processes, and equipment, and connects those models to experiments and physical manufacturing through AI. Semiconductor manufacturing is our primary proving ground, while functional materials broaden the design space.
Multiscale Computational Science
Many engineering systems are governed by phenomena that cross length, time, and physical domains. We develop first-principles, atomistic, mesoscale, feature-scale, and continuum models, then connect them with physics-based and data-driven scale bridging. The result is not a collection of isolated simulations but a predictive description that carries mechanisms and uncertainty across scales.
More detail →Virtual Design of Materials, Processes & Equipment
We turn computational understanding into design variables and decisions. Our work spans semiconductor surfaces and interfaces, etching and deposition, film growth, reactor transport, and equipment conditions, with AI and optimization used to explore spaces that are too large for trial and error. The same methodology extends to broader functional materials.
More detail →Advanced Manufacturing AI
Advanced Manufacturing AI begins when virtual models and physical systems share information. We combine simulation, experimental databases, in-situ sensing, and reduced-order models in digital twins that can estimate hidden process states, explain predictions, and support optimization and control. Physical AI and agentic systems are the path toward closing this loop from observation to action.
More detail →Active projects
AI-Based Integrated Virtual–Physical Platform for Ultra-Precision Cryogenic Etching
Couples virtual process simulation with physical measurement to model and predict ultra-precision cryogenic etching.
Integrated Virtual–Physical AI for Ultra-Precision Manufacturing of Advanced Semiconductor Devices
A virtual–physical AI digital twin for the fabrication processes of advanced semiconductor devices.
반도체 웨이퍼 나노 구조체 증착 공정의 시뮬레이션 기반 해석 및 예측 플랫폼 구축
Industry-commissioned platform for simulation-based analysis and prediction of nanostructure deposition processes.
원자층 공정 모사 시뮬레이션 및 머신러닝 기반 공정 예측 기술개발
Atomistic MD and DFT simulation of ALD and ALE surface reactions as the physical basis for machine-learning process prediction in ETRI's intelligent atomic-layer process program.
MLFF 기반 Autonomous Reaction Map과 CFD 연계를 통한 저온 MoS₂ CVD 기상 반응 해석
MLFF-driven autonomous reaction mapping coupled with CFD for low-temperature MoS₂ CVD gas-phase chemistry.
Advancing Machine Learning Force Fields for Semiconductor Processing with NVIDIA Technologies
GPU-accelerated machine-learning force fields for semiconductor process modeling.
Area Selective Deposition of Novel Metals with 100% Selectivity for Interconnect Technology of Si Devices
Area-selective deposition chemistry and processes for interconnect and electrode technology.
Development of Oxide Semiconductor Materials and Process Technologies for Next-Generation DRAM Applications
Oxide-semiconductor materials and processes for next-generation DRAM devices.
Low-Temperature Chalcogenide Semiconductor Materials with High Mobility for Large-Area 3D Integration
Low-temperature chalcogenide semiconductors with high mobility for large-area 3D integration.
차세대 모빌리티향 센서-뉴로모픽 반도체 지능형 통합시스템 개발
Intelligent integrated sensor–neuromorphic semiconductor systems for next-generation mobility.
반도체특성화대학지원사업
Semiconductor-specialized undergraduate education program at UNIST.
반도체특성화대학원지원사업
Semiconductor-specialized graduate school program (GS-SMDE) at UNIST.
KISTI Innovation Support Program (Supercomputing)
National supercomputing allocation supporting large-scale simulations.
Completed projects
Development of Novel Materials for Cryogenic Etching Process
KISTI Innovation Support Program (Supercomputing)
UNIST Supercomputing Center Researcher Support Program
Funded & supported by