On December 23, 2025, NC AI and Korean CAD firm ZYX Technology signed a strategic MOU to jointly develop industrial digital twins using 3D generative AI. The partners plan to combine NC AI’s multimodal 3D generation model VARCO 3D with ZYX’s CAD and digital-twin platforms for manufacturing, construction and logistics use cases.
This article aggregates reporting from 2 news sources. The TL;DR is AI-generated from original reporting. Race to AGI's analysis provides editorial context on implications for AGI development.
This MOU is a good snapshot of where ‘physical AI’ is headed. NC AI brings a multimodal 3D generative model (VARCO 3D) capable of turning text and images into realistic 3D assets; ZYX brings industrial‑grade CAD and digital‑twin pipelines already used in smart construction and manufacturing. Together they’re aiming to build rich, simulation‑ready digital twins that obey physical constraints closely enough to train robots, planners and industrial control systems.
For the AGI race, this is part of a broader trend: world models are moving from research demos into production infrastructure. High‑fidelity 3D generative models coupled with live CAD and sensor data let AI systems experiment in silico, then push actions back into the real world. That tight virtual‑physical loop is a prerequisite for more general, embodied intelligence. Korean players are also signaling they don’t intend to cede this domain to US or Chinese incumbents.
Competitively, NC AI is extending beyond gaming (NCSoft’s heritage) into cross‑industry platforms, while ZYX gets access to advanced generative tech that can make its CAD tools far more autonomous. If they can show credible productivity gains in factories and construction sites, expect other CAD and PLM vendors globally to pursue similar alliances with multimodal model shops.



