Network as a Data Center: Samsung and NVIDIA to Develop Next-Generation Mobile Networks with Integrated AI

Samsung Electronics announced a collaboration with NVIDIA to advance AI-RAN technologies to accelerate the adoption of AI in mobile networks by expanding the CPU ecosystem and strengthening partnerships with GPU developers.

Samsung has announced significant progress in bringing AI to its O-RAN radio access networks. One of the most important achievements was the achievement of interoperability between Samsung vRAN and NVIDIA accelerators. The latter’s seamless integration with software-defined networks allows for expanded AI capabilities in networks, all on the basis of common commercial servers.

The companies will also continue to work together to integrate Samsung’s vRAN with NVIDIA Grace Arm processors and NVIDIA accelerators (using CUDA). The new solutions, according to the companies, are suitable for deployment anywhere – from rural and suburban areas to densely populated cities. As a founding member of the AI-RAN Alliance consortium, established in 2024, Samsung is actively involved in advancing AI-RAN technologies together with academic institutions and leading manufacturers.

Image source: Samsung

Samsung said its end-to-end software-defined network architecture is the optimal foundation for easily deploying and implementing AI at every layer of the network. This paves the way for network infrastructure to be used for both mobile communications and other workloads. The company noted that this can enable a data center-like network architecture that will unlock new business opportunities.