NTT to Triple Data Centre Capacity as Part of AI-Native Network Plan

Why It Matters

The expansion positions NTT as a primary AI‑ready infrastructure provider in Japan, enabling enterprises to scale generative‑AI workloads locally and reducing reliance on overseas cloud services. It also signals a broader shift toward dedicated AI data‑centres, reshaping the competitive landscape of the global data‑centre market.

Key Takeaways

• NTT aims for 1 GW data‑centre capacity by 2033
• Current capacity 300 MW; new sites add 350 MW by 2029
• AIOWN targets GPU‑heavy inference workloads for enterprises
• 2024 capex $13.1 bn; data‑centre spend $2.7 bn
• Containerized centres enable rapid, customizable AI compute deployment

Pulse Analysis

NTT’s aggressive push to reach 1 GW of data‑centre capacity reflects the accelerating demand for AI‑specific infrastructure. While traditional cloud providers have long dominated the market, the shift from training‑centric to inference‑centric AI workloads requires low‑latency, power‑dense environments that can be colocated with enterprise networks. By integrating GPU‑rich servers, upgraded backbone connectivity, and advanced cooling, NTT’s AIOWN strategy promises an "optimal usage environment" that reduces latency and operational costs for Japanese firms deploying generative AI across sectors such as manufacturing, automotive and robotics.

The financial backdrop underscores the scale of the opportunity. McKinsey projects $5.2 trillion in global AI‑focused data‑centre spend through 2030, and NTT’s 2024 capex of $13.1 bn—of which roughly $2.7 bn was earmarked for data‑centre development—suggests the company is gearing up to capture a sizable share of the domestic market. The upcoming 100 MW Tochigi facility and the 250 MW Inzai/Shiroi centre will not only boost total capacity but also diversify geographic coverage, enhancing resilience and reducing latency for customers spread across Japan’s 47 prefectures.

Beyond raw capacity, NTT’s introduction of container‑type, modular data centres offers a flexible deployment model that can be rapidly scaled to meet fluctuating AI demand. This approach mirrors trends seen in Western markets, where hyperscale operators are experimenting with portable, edge‑focused solutions. For Japanese enterprises, the ability to customize location, size and equipment could accelerate AI adoption in regulated industries that require data sovereignty. As NTT’s upcoming earnings release may reveal higher-than‑expected capex forecasts, the company is poised to become a cornerstone of Japan’s AI infrastructure ecosystem, challenging both domestic rivals and global cloud giants for market share.