SoftBank Corp reportedly outlined plans to invest $138 million in advanced computing infrastructure to power its own generative AI model, aimed at corporate clients.
News outlet Nikkei Asia reported the company will build infrastructure equivalent to a supercomputer, which will be powered by Nvidia’s graphics processing unit for large language models.
A computing infrastructure of this kind is indispensable for developing cutting-edge AI systems, Nikkei Asia stated.
As part of the process to develop the offering, the operator will target call centre operators as it hopes to build generative AI tools tailored to various enterprise clients, including finance and healthcare.
It is also planning to lease its new computing infrastructure to businesses developing their own AI models.
Earlier this year, SoftBank Corp president and CEO Junichi Miyakawa announced plans to power data centres in the country with Nvidia’s latest AI chip, claiming the facilities will be capable of hosting generative AI, 5G and future 6G services.
Miyakawa claimed the demand for data processing will rapidly increase at a time when “society co-exists with AI”.
SoftBank also deployed an AI chat service for its 20,000 employees.