Foxconn is quietly using Sharp as a local front door to push $NVDA GPU servers into Japan's enterprise market.

Sharp-branded 8-GPU RTX Pro boxes. Built by Foxconn. Aimed at on-prem inference, not hyperscale training.

Target: ¥250 billion by FY2030.

Smart play. Sharp brings the brand recognition and service network in Japan. Foxconn brings the manufacturing scale. $NVDA gets enterprise distribution without building it from scratch.

This is how AI infrastructure gets monetized outside the hyperscalers — localized, packaged, and sold through trusted regional brands.