Samsung Electronics and NVIDIA Corporation are expanding their deep-rooted 25-plus year semiconductor collaboration to include foundry services, manufacturing, AI, and robotics. Together they are innovating breakthroughs across next-generation semiconductors and powering Samsung’s devices, services, and robotics ecosystems. Samsung is accelerating its lithography platform for advanced chip-making using NVIDIA CUDA-accelerated infrastructure, achieving an approximately 20× performance gain in computational lithography and TCAD simulations. Through digital twin technology via NVIDIA Omniverse, Samsung’s global fabs will shorten design-to-operation time and enable AI-driven predictive maintenance, operational optimisation and real-time decision-making. Samsung is leveraging NVIDIA GPUs, CUDA-X libraries, and partner tools from Synopsys, Cadence and Siemens to accelerate simulation, verification and manufacturing analytics.
AI Factory Collaboration: A New Era for Manufacturing
At the APEC Summit, NVIDIA announced a major joint initiative with Samsung: the construction of a new AI factory which marks an intersection of intelligent computing and advanced chip production.
This factory, powered by more than 50,000 NVIDIA GPUs, is positioned to become the cornerstone of Samsung’s digital transformation, embedding accelerated computing directly into full-scale advanced chip manufacturing.
The collaboration sets a new global standard for AI-driven semiconductor manufacturing, integrating data streams from physical equipment and production workflows to enable predictive maintenance, process improvements and autonomous fab operations.
Strategic Alliance – History and Future Scope
Samsung and NVIDIA’s partnership spans over 25 years, tracing back to collaborations such as NVIDIA’s NV1 graphics card using Samsung DRAM, and progressing through industry-first commercial HBM (high bandwidth memory) and current supply arrangements for HBM3E and HBM4.
The companies plan to extend their alliance beyond memory and high-density modules into broader custom solutions, system-on-chip (“SoC/AMM”) designs and foundry services supporting the wider semiconductor ecosystem.
Also Read: SymphonyAI Launches IRIS Forge to Boost Industrial AI Platform
Technology Highlights: Accelerated Tools & Digital Twin Factory
Samsung is deploying NVIDIA Omniverse to build digital twins of its global fabs. These virtual environments provide physically-accurate simulations, enabling Samsung to reduce time from design to operations and leverage data for AI-driven factory automation and real-time decision-making.
To manage intelligent logistics, Samsung is using NVIDIA RTX PRO Servers with NVIDIA RTX PRO 6000 Blackwell GPUs. This supports a real-time digital twin of the fab, enabling operational planning, anomaly detection and logistics optimisation key steps toward a fully autonomous manufacturing facility.
In the most computationally intensive task of chip manufacturing computational lithography Samsung and NVIDIA are integrating the NVIDIA cuLitho library into Samsung’s OPC lithography platform. Their collaboration has yielded roughly 20× greater performance and scalable deployment across semiconductor manufacturing.
Samsung’s smart manufacturing strategy also encompasses robotics and generative AI: The company builds proprietary AI models supporting more than 400 million Samsung devices, delivering real-time translation, multilingual interactions and intelligent summarisation.
In robotics, Samsung is leveraging NVIDIA Isaac Sim (on Omniverse) plus NVIDIA Cosmos world foundation models and the NVIDIA Jetson Thor edge AI platform, aiming to deploy robots that understand and interact with the physical world in real time.
Broader Ecosystem & Network Integration
As part of the collaboration, Samsung and NVIDIA are working with Korean telecom operators and academic partners to develop AI-RAN network technology. This brings together AI and mobile network workloads a necessary step for widespread adoption of physical AI applications.
