Samsung and NVIDIA are significantly strengthening their 25-year collaboration in a global agreement that provides for the deployment of 50.000 GPUs of the green giant in the AI megafactory of the South Korean giant and will work in various areas such as HBM4, EDA tools and AI-RAN.
Samsung’s AI megafactory and the supply of the 50,000 accelerators is the most significant part of the announcement. Samsung intends to use this AI megafactory to integrate its labyrinthine manufacturing processes into “a single smart grid, where AI continuously analyzes, predicts and optimizes production environments in real time”.
To expand your AI factory in the coming yearsSamsung also plans to take advantage of NVIDIA accelerated computing and implement digital twin manufacturing with the help of NVIDIA Omniverse libraries, the simulation platform for creating and operating applications, services and 3D virtual worlds in real time.
In addition, manufacturing using digital twins allows you to create a virtual replica of a physical manufacturing processwhich allows companies to monitor and analyze the entire process in real time. Samsung plans to build digital twins to digitally visualize its manufacturing operations, using these “virtual environments to identify anomalies, perform predictive maintenance, and optimize production before changes are applied to the physical world”.
Over time, Samsung plans to expand its AI factory infrastructure to its global manufacturing centers, including its US base in Taylor.
Samsung and NVIDIA, more agreements
Samsung and NVIDIA have announced other types of collaborations and the most important is ‘HBM4 supply‘. With high bandwidth and energy efficiency, Samsung’s HBM solutions are expected to help accelerate the development of future AI applications and lay the fundamental foundation for manufacturing infrastructure powered by these technologies.
Built with the company’s 10 nanometer (nm) sixth-generation DRAM memory and 4 nm logic base chip, Samsung HBM4’s processing speed can reach 11 gigabits per second (Gbps), the fastest in the industry, far exceeding the JEDEC standard of 8 Gbps.
Samsung and NVIDIA are also collaborating with other electronic design automation (EDA) partners to develop next-generation GPU-accelerated design tools and technologies. EDA is used to automate the design, verification, and preparation for manufacturing of integrated circuits and microprocessors.
The two companies are also collaborating in the field of intelligent robotics through the NVIDIA RTX PRO 6000 Blackwell Server Edition platform and the NVIDIA Jetson Thor robotics platform.
Finally, Samsung is working with NVIDIA, Korean telecom operators, academia and research institutions to foster collaboration in AI-RANa next-generation communications technology that integrates AI into cellular radio access networks to improve their performance, efficiency and capabilities.
