NTT Data Group, owned by government-backed NTT and Japan's largest data center operator, plans to invest at least $9 billion through 2033 to quadruple its computing capacity to 1 gigawatt by adding roughly 750 megawatts over the next seven years.
The expansion reflects surging demand from Japanese companies seeking to close the gap with global peers in artificial intelligence infrastructure.
What happened
NTT Data Group, Japan's largest data center operator, plans to spend at least $9 billion through 2033 to add roughly 750 megawatts of capacity over the next seven years, quadrupling its total computing capacity to 1 gigawatt.
Why it matters
Japanese companies are racing to build AI infrastructure to keep pace with global competitors. NTT Data's expansion—the country's biggest data center operator moving aggressively on capacity—signals the scale of domestic demand for AI computing resources, particularly from enterprises catching up in the technology.
What to watch
The buildout runs through 2033. The cost estimate of at least ¥1.5 trillion ($9.6 billion) assumes ¥2 billion to ¥2.5 billion per megawatt for construction and equipment, excluding the price of AI accelerators like Nvidia's hardware.
NTT Data Group, the country's largest data center operator and a subsidiary of government-backed NTT, is preparing a major expansion to meet explosive demand for AI computing capacity among Japanese enterprises. According to people familiar with the company's plans, NTT Data intends to add roughly 750 megawatts of capacity over the next seven years—a buildout running through 2033—with the goal of quadrupling total computing capacity to 1 gigawatt.
The investment required for this expansion is substantial. NTT Data expects to spend at least $9 billion (¥1.5 trillion) on construction and equipment installations. This figure is based on an industry assumption that each megawatt of data center capacity installed in Japan costs between ¥2 billion and ¥2.5 billion, though the estimate excludes the cost of AI accelerators such as Nvidia's specialized processors. The exclusion of accelerator costs suggests these expenses will be significant and tracked separately in the company's budgeting.
The expansion is being driven by a fundamental shift in corporate strategy across Japan. Companies are racing to build and access AI infrastructure to close a perceived gap with international competitors. By becoming the primary infrastructure provider for this wave of domestic AI adoption, NTT Data is betting that its position as Japan's biggest data center operator—combined with government backing through its parent company—will make it the natural choice for enterprises seeking reliable, large-scale computing resources. The seven-year timeline and multi-billion-dollar commitment represent a long-term wager on the durability of Japan's AI demand.
NTT Data's planned expansion reflects a broader shift in Japan's technology landscape. As companies across the country race to adopt artificial intelligence, the infrastructure to support that transition has become a critical bottleneck. By announcing this commitment—one of the largest data center buildouts in Japan—NTT Data is positioning itself to serve the surge in domestic demand for computing capacity. The timeline through 2033 and the scale of investment (at least $9 billion) underscore how serious the commitment is.
The cost structure the article describes—¥2 billion to ¥2.5 billion per megawatt for construction and equipment—provides a concrete benchmark for what data center infrastructure costs in Japan, excluding the price of specialized AI chips like Nvidia's processors. This suggests NTT Data is planning for a fully built-out, modern facility, not a low-cost retrofit. The decision to keep Nvidia hardware costs separate in the estimate hints at the scale of chip purchases the company will need to make to equip these centers for AI workloads.
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytime
Ask AI anything about this article. Q&As are published on this page for other readers too.
The AI news that matters, in one minute each morning.
Sign up free