NTT DATA and NVIDIA: AI Factories to Revolutionize Business

Artificial intelligence (AI) initiatives often face significant obstacles when being implemented in businesses. Although these projects can be impressive during demonstrations, they struggle to evolve into effective operational tools. To overcome these challenges, NTT DATA and NVIDIA have decided to collaborate to offer an innovative solution.

A Collaboration to Accelerate AI Adoption

NTT DATA and NVIDIA have launched a joint initiative aimed at accelerating the deployment of AI in the business world. Their project is based on the creation of "AI factories," platforms designed around NVIDIA's advanced technologies. These AI factories combine the computing power of NVIDIA's GPUs with high-performance network connectivity, integrating enterprise AI software such as NVIDIA NeMo and NVIDIA NIM microservices.

These platforms are designed to support businesses throughout the AI lifecycle, from model training to application development, while adhering to strict governance rules. The goal is to standardize outcomes and reduce the time and costs required to transform a promising pilot project into a viable operational solution.

NeMo and NIM: The Pillars of the Platform

On the technical side, the architecture of these factories relies on two key software components developed by NVIDIA. The first, NeMo, is a suite dedicated to creating multi-agent AI systems on GPU-accelerated infrastructures. The second, NIM microservices, offers preconfigured and GPU-optimized containers, accompanied by APIs that facilitate the deployment of artificial intelligence applications.

Together, these technologies form a comprehensive AI agent platform, ready for production. NTT DATA also offers pre-qualified GenAI prototypes built on this architecture, aimed at reducing technical complexity and accelerating return on investment for companies developing industry-specific applications.

According to John Fanelli, companies are now looking for platforms capable of supporting their AI initiatives beyond the mere testing phase. The solutions offered by NTT DATA provide clients with scalable environments tailored to their business needs.

Concrete Applications Across Various Sectors

The AI factories of NTT DATA and NVIDIA are already finding concrete applications in several sectors. In the medical field, for example, a cancer research center is using NVIDIA HGX platforms, in collaboration with NTT DATA and Dell Technologies, to conduct advanced radiological analyses and accelerate the evaluation of models supporting clinical research.

In the automotive industry, a global supplier is using an AI-based factory architecture and NVIDIA infrastructure to validate its workloads directly on the hardware before large-scale deployment, significantly reducing production lead times.

In the technology sector, an American company is leveraging NVIDIA's accelerated simulation and 3D visualization technologies to test and optimize a next-generation battery production line, validating the operation of the facility in a digital environment before its physical construction.

These examples demonstrate that enterprise AI factories can be adapted to different sectors, with NVIDIA's software suite serving as a common infrastructure, while applications are customized according to the specific needs of each industry.

Challenges to Overcome for AI Factories

Although promising, the AI factories of NTT DATA and NVIDIA will need to overcome several obstacles to achieve successful large-scale deployment. The first challenge concerns infrastructure. The platforms rely on very powerful GPU architectures, which are costly and energy-intensive. For many organizations, installing or renting this type of environment represents a significant investment, especially when AI projects do not yet guarantee an immediate return.

Technical complexity is also a barrier. Integrating AI into existing IT systems remains a delicate task, and many fail in this endeavor. A report from the MIT State of AI in Business 2025 indicates that 95% of AI pilot projects remain stalled before delivering real business value, often due to models that are ill-suited to internal processes and existing systems.

Another challenge relates to data governance and regulatory compliance. AI models often handle sensitive information, and in sectors like healthcare or industry, organizations must ensure security, traceability, and compliance with local regulations.

Finally, the issue of standardization remains central. Each sector has its own business constraints, and transforming a generic platform into a solution truly tailored to a specific field still requires a great deal of customization and expertise.