NTT Global Data Centers wants to double data center capacity to 4 GW as AI demand booms

NTT Global Data Centers plans to double the capacity of its global data center network to 4 GW to keep pace with the surge in demand for AI infrastructure. It is one of the most notable signals of spring 2026: the race for artificial intelligence is increasingly constrained not only by chips, but by electricity, sites and cooling.

The scale of the expansion

NTT's data center division, considered the third-largest data center operator outside China, wants to reach 4 GW of capacity within about two years. The company currently has more than 2 GW of critical IT load, over 160 facilities and a presence in more than 20 countries and regions.

To double in size, NTT is running 34 projects at once. This is not about launching a couple of new sites here and there, but a large cycle of construction and expansion across several regions simultaneously.

The readiness of demand matters separately. According to the head of the business, Doug Adams, contracts are already secured for more than 70% of these 34 projects. For the market this is a strong marker: NTT is not building speculatively, but against customer needs that are already confirmed.

Moreover, the company expects to be able to offer significantly more than 5 GW within a five-year horizon. If this plan comes through, NTT will secure its place among the players that profit from the AI boom not at the level of models, but at the level of physical infrastructure.

Who is taking the capacity

Demand comes not only from traditional cloud companies. In March, NTT reported new commitments of almost 115 MW across campuses in Virginia, Illinois and California. Of that volume, more than 90 MW was reserved by a large hyperscaler, and almost 20 MW more was contracted by three enterprise customers.

It is a good indicator of the market's structure: the main consumers are still the cloud giants, but large enterprises are also starting to claim capacity early for their future AI workloads.

Customers need more than square meters for racks. They need sites designed from the outset for dense computing and fast growth. In NTT's materials, this looks like this:

• support for high-density compute workloads for AI services
• the option to move to liquid cooling as requirements grow
• proximity to cloud availability zones and low latency
• compliance with regulatory and security requirements
• fast capacity delivery and a clear path to further expansion

Over the past year, NTT commissioned ten new facilities in North America, EMEA and Asia-Pacific, adding more than 370 MW of IT capacity. This shows the company does not merely announce ambitious targets, but can quickly convert demand into real infrastructure.

Amid the generative AI boom, this becomes a competitive advantage in its own right: large customers increasingly choose not the cheapest operator, but the one that can deliver capacity on time and without compromises on cooling, networking and fault tolerance.

Money and execution

The next question is how to finance such growth. NTT says it can use its own funds, partner capital and the tools of its REIT, NTT DC REIT. This is an important detail, because the data center market is now growing faster than many developers can secure electricity, equipment and contractors. Having several funding channels lowers the risk of projects stalling midway, and contracts signed in advance make the investment case for NTT noticeably more convincing.

But money is only part of the task. In practice, the pace of bringing new AI-ready sites online depends on access to power grids, lead times for electrical and mechanical equipment, high-density design and the preparation of cooling systems. So the NTT announcement matters not only for the 4 GW figure. It shows that large operators have learned to plan development years ahead for future demand and to sell capacity before a facility is fully ready.

For customers this means one thing: free capacity in key markets is shrinking, and it has to be reserved earlier than before.

What this means

The winners of the AI race in 2026 are determined not only in labs and not only in the GPU market. Operators that can quickly turn electricity, land and capital into finished data centers are gaining importance. NTT's 4 GW plan shows that the infrastructure layer of the AI market is becoming as strategic as the models and cloud services themselves.