Sakura Internet announced it could increase its fiscal 2026 capital spending nearly seven times from the initial ¥4.4 billion plan to between ¥20 billion and ¥30 billion (about $125 million to $190 million) to meet growing AI demand in Japan [1, 2, 3]. CEO Kunihiro Tanaka said AI server utilization rates are currently at 80% to 90%, requiring significant capacity expansion and new AI accelerators including GPUs [1, 2]. "AI server usage rates are 80% to 90%. We need to expand capacity significantly because the number of orders are increasing sharply," Tanaka said [2].
The company has already invested roughly half of its planned ¥113 billion capital expenditure for 2023-2030 on AI hardware such as Nvidia GPUs [1]. The Japanese government pledged subsidies of up to ¥57.5 billion to cover about half of Sakura Internet’s data center spending, supporting the investment [1]. Sakura Internet is the only Japanese data center operator approved for cloud service use by the Japanese government and local municipalities alongside US firms Amazon, Google, Microsoft, and Oracle [1].
Tanaka noted hopes to procure new GPUs within fiscal 2026, investing up to ¥30 billion for capacity expansion to handle sharply increasing customer orders [3]. Sakura Internet initially announced in April 2026 its capital spending plan of ¥4.4 billion for the fiscal year but revised the outlook following surging AI workloads [3] [1, 2].
The company is focusing on scaling AI infrastructure to meet client demand, emphasizing accelerated procurement and deployment of new hardware layers. The revised investment plan marks a major escalation in Sakura Internet’s commitment to AI capacity for fiscal year 2026 [1, 2].
