Fujitsu networking arm in Nvidia virtual RAN team up - SDxCentral

Fujitsu's entry into Nvidia's vRAN ecosystem is not a vendor partnership—it is a generational shift in the RAN compute substrate. Baseband processing is migrating from dedicated DSPs to general-purpose GPU acceleration, and once this architecture locks in at operator level, the moats protecting Qualcomm's RAN division and MediaTek's baseband units will be structurally eroded within two to three product cycles. The geopolitical signal is equally pointed: a US-Japan trusted-supplier corridor in critical telecom infrastructure mirrors the post-2018 bifurcation driven by US export controls on advanced nodes, systematically excluding Chinese vendors from standards-setting. On the competitive front, Ericsson and Nokia will accelerate O-RAN disaggregation to avoid GPU lock-in, while Samsung will leverage its AI-RAN narrative to contest positioning. The O-RAN Alliance will be forced to introduce compute-agnostic interface specifications. Over the next 12 to 24 months, expect GPU-native 6G RAN prototypes entering operator trials, a consolidation wave among mid-tier RAN vendors, and a measurable structural shift in Nvidia's data center revenue mix toward telecom workloads.