NTT DATA says it has opened AI Factory sites in Paris and Singapore, adding physical facilities where organisations can test AI workloads before committing to deployment. The expansion gives enterprise teams new places to assess infrastructure, workload placement and operating costs across a network spanning three continents.
Watch Desk analysis
What happened
The new sites join NTT DATA facilities in Plano, Texas, and Riyadh, as well as its Munich AI Factory, which opened on 7 October. The company describes the network as a way for clients to run discovery workshops, technical deep dives, proof-of-concept planning and solution design. Its announcement says the facilities support use cases including employee productivity, customer experience, cybersecurity, networking and software development.
The factories combine NTT DATA’s services with Cisco and NVIDIA technology, including the Cisco Secure AI Factory with NVIDIA infrastructure platform. NTT DATA says clients can test systems they already own and control in a production-grade environment before making deployment decisions.
Why it matters
AI infrastructure is expensive enough that testing a workload against real equipment, security and operating requirements before scaling it has obvious appeal. These sites offer a structured route from an initial idea to a proof of concept, with physical access points across Europe, Asia and North America.
That is a useful offer, but not evidence that a particular AI project will become cheaper or succeed in production. The value will depend on what clients can actually test, how representative the facilities are of their eventual deployments, and whether the work leads to sound decisions rather than a very polished tour.
Our read
The important development is not simply another AI infrastructure partnership. NTT DATA is extending a network designed to let organisations test workloads before they spend heavily on deployment. That could help turn vague interest into a more concrete technical and cost assessment. The proof will be in what customers learn and whether those trials lead to working systems, not in how many cities can be added to a map.
What to watch
- What workloads and infrastructure configurations clients can test at each site.
- Whether NTT DATA publishes examples of trials that progress into production.
- How the facilities help customers assess costs, security and workload placement.
Discussion spark: Should companies test AI infrastructure in shared facilities before investing, or does a useful trial need to happen on their own systems?
Sources and evidence
- NTT DATA Expands AI Factory Footprint with Cisco and NVIDIA – Data Centre Magazine (28 October 2026, 15:05 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.