MSI has announced it is showcasing an AI virtual Radio Access Network (AI-vRAN) stack integrated with GPU server systems at Mobile World Congress (MWC) 2026 (Booth 5A61). MSI says the goal is to bring AI capabilities into core network operations to support growing telecom workloads and speed the transition toward AI-powered mobile networks.

MSI positions the platform as a unified architecture intended to simplify deployments across Open Radio Access Network (O-RAN), private 5G, and virtual RAN environments. MSI says this approach is designed to keep network functions consistent across distributed and centralized infrastructure while enabling adoption of AI-powered network functions.

At MWC 2026, MSI is highlighting two GPU server platforms for different AI-vRAN deployment profiles: the CG480-S6053 and CG290-S3063. MSI says the CG480-S6053 supports dense GPU configurations for compute-intensive AI inference and acceleration workloads, while the CG290-S3063 targets space- and power-efficient deployments in a 2U chassis for edge and distributed network environments. MSI says both systems support flexible configurations of GPUs, network interface cards (NICs), data processing units (DPUs), and storage to tune performance, scalability, and cost efficiency across a consistent infrastructure. MSI also lists the 2U CX271-S4056 platform (HE SKU) for “balanced data processing for AI workloads,” aimed at scenarios where compute efficiency and system balance are critical.

MSI says the AI-vRAN solution supports dynamic GPU allocation, shifting resources between 5G communication and AI workloads based on real-time demand. MSI also says the platform can run Base Station and Edge AI functions on the same infrastructure, enabling simultaneous processing of RAN and AI workloads while improving compute utilization, with integrated software intended to support deployment and operation of AI services at the network edge.

For scaling, MSI says it offers GPU-accelerated server configurations ranging from two to eight GPUs for deployments spanning wireless access, metro edge, core network, and centralized cloud environments. MSI also says its AI platforms are built on the NVIDIA MGX architecture and that this approach has been widely adopted in AI data centers; MSI says it is extending the same data center-class computing principles to telecom to meet next-generation mobile-network requirements for performance, scalability, and efficiency.

Source: MSI