NVIDIA and Samsung have decided to build a semiconductor AI factory that will be powered by more than 50,000 NVIDIA GPUs. The aim is to establish an AI-driven production foundation by combining Samsung’s semiconductor technologies with NVIDIA platforms. The hope of the two firms is to accelerate agentic and physical AI applications for advanced chip manufacturing, mobile devices, and robotics.
The companies reiterated that they have a longstanding alliance, having collaborated for more than 25 years, starting with the DRAM for NVIDIA’s first graphics card, and continuing through key supply for HBM3E and HBM4. With today’s news, the partnership is expanding from memory solutions and custom solutions to foundry services for manufacturing, AI, and robotics.
“We are at the dawn of the AI industrial revolution — a new era that will redefine how the world designs, builds and manufactures,” said Jensen Huang, founder and CEO of NVIDIA. “As Korea’s and one of the world’s foremost technology and industrial leaders, Samsung is forging its AI foundation with NVIDIA to lead the future of intelligent and autonomous manufacturing — transforming Samsung itself and the many industries around the world built on Samsung technologies.”
“NVIDIA has been a visionary of this new AI era, and its technologies have empowered innovators to reinvent industries,” said Jay Y. Lee, executive chairman of Samsung Electronics. “From Samsung’s DRAM for NVIDIA’s game-changing graphics card in 1995 to our new AI factory, we are thrilled to continue our longstanding journey with NVIDIA in leading this transformation as we envision creating new standards for the future and accelerating breakthroughs for the world.”
Using the NVIDIA in the AI factory, Samsung will be able to make several key technological accelerations. First, in computational lithography, The NVIDIA cuLitho library will be integrated into Samsung’s OPC platform, giving it a 20x performance gain. Another area where they’re helping each other is digital twins; Samsung is using the NVIDIA Omniverse platform to build physically accurate digital twins of global fabs to enable shorter time from design to operations, AI-driven predictive maintenance and operational optimization, and real-time decision-making.
Samsung also uses NVIDIA GPUs for its chip design and verification, specifically it uses CUDA-X libraries with solutions from Synopsys, Cadence, and Siemens for big speedups in circuit simulation, verification, and manufacturing analysis. Samsung is also developing intelligent robotics for manufacturing automation and humanoid applications using NVIDIA RTX PRO Servers, NVIDIA Isaac Sim, and NVIDIA Jetson Thor. Finally, the factory will accelerate intelligent logistics and operation planning.