SoftBank is making sovereign GPU-powered neocloud infrastructure generally available in Japan starting in October, following a beta launch this week. The new infrastructure runs on a combination of Nvidia GB200 NVL72 GPUs and its home-grown Infrinia AI Cloud OS that provides both Kubernetes as a Service and Inference as a Service capabilities.
Infrinia was unveiled in January.
"Under our new growth strategy, 'Activate AI for Society,' SoftBank will provide integrated computing infrastructure and software that can be securely used within Japan as a neocloud provider," Junichi Miyakawa, President & CEO of SoftBank Corp, said in a statement.
SoftBank said it is aiming to optimize AI processing from training to inference by integrating its AI Data Center GPU Cloud with AI-RAN, while building a sovereign, distributed AI infrastructure that delivers low latency and high reliability.
The move reflects a growing trend among operators launching sovereign cloud and neocloud operations. Other operators who have done the same include Norway's Telenor, Germany's Deutsche Telekom, South Korea's SK Telecom, Canada's Telus and France's Orange.
During its beta period, SoftBank said it plans to use its sovereign cloud infrastructure internally across its business units.
Read the full SoftBank press release here.