VIDIA Partners Marvell to Boost AI-RAN, Ecosystem Gains Momentum

According to industry reports, NVIDIA has formed a strategic alliance with Marvell, including a US$2 billion investment, to accelerate development in silicon photonics and expand its AI-RAN ecosystem. Beyond optical interconnect innovation, NVIDIA aims to integrate Marvell into its broader vision of transforming global telecommunications networks into AI-driven infrastructure.

NVIDIA previously initiated its AI-RAN strategy through an investment in Nokia and is now strengthening this initiative by collaborating with Marvell. The partnership will leverage NVIDIA’s Aerial AI-RAN platform for 5G and 6G, enabling telecom networks to support AI workloads through advanced optical interconnects and silicon photonics technologies.

AI-RAN (Artificial Intelligence Radio Access Network) enhances base station capabilities by incorporating GPU-based computing. It operates across three key layers: real-time optimization of network resources to improve performance; transforming base stations into server-like architectures capable of handling both network traffic and AI inference tasks; and enabling edge computing, where each base station acts as a low-latency node delivering AI services directly to end devices such as autonomous vehicles, smartphones, and robotics systems.

Industry observers note that expanding the AI-RAN ecosystem could significantly broaden AI deployment opportunities. By embedding AI capabilities into telecom infrastructure, NVIDIA is effectively creating a new growth engine for the AI economy.

Ecosystem partners are also advancing related solutions. Lite-On Technology has introduced AI-RAN offerings based on NVIDIA’s Aerial platform, focusing on improved compatibility, simplified integration, and scalable deployment. Pegatron has outlined its AI-RAN strategy by integrating intelligent computing into O-RAN architectures, supported by proprietary interference suppression technologies to ensure stable, low-latency communication in complex environments such as industrial and healthcare settings. Wistron NeWeb has also collaborated within the ecosystem to accelerate innovation in AI-native 5G-Advanced and 6G, including the development of AI-enabled open radio unit (O-RU) solutions.

Overall, the collaboration between NVIDIA and Marvell highlights a growing convergence between AI computing and telecommunications infrastructure, positioning AI-RAN as a key enabler of next-generation network evolution and semiconductor demand growth.