Signals AI data center push into Europe with France's Bull

1 June 2026, Taipei, Taiwan – Driving next generation artificial intelligence infrastructure and industrial transformation, Hon Hai Technology Group (Foxconn) (2317) is demonstrating its growing capabilities as a technology manufacturing platform service provider this week at COMPUTEX 2026, with its cutting-edge technology products supporting NVIDIA’s Vera Rubin super-computing platforms, and advances in data center build-out, robotics, healthcare and space.

At the same time, Foxconn signaled its readiness to bring comprehensive AI data center solutions to Europe in a new strategic partnership with AI high-performance computing leader Bull. The French company will focus on system design and Foxconn will lead manufacturing. By combining technological expertise and supply chain strengths, the collaboration aims to accelerate the development of AI infrastructure in Europe.

Amid strong global growth in the AI sector, Foxconn continues to deepen its AI server deployment and has become an indispensable partner within the global AI supply chain. The world’s largest electronics manufacturer is advancing from an AI server supplier to a “Token Factory” provider, further extending into full AI ecosystem integration.

AI Infrastructure

At COMPUTEX Booth M0120, Foxconn is showcasing its manufacturing and system integration capabilities for the NVIDIA Vera Rubin NVL72, along with the latest AI server and high-speed interconnect technologies, including:

• NVIDIA Groq 3 LPX
• NVIDIA Vera CPU platform
• NVIDIA HGX Rubin NVL8 platform
• NVIDIA MGX 2U and 4U modular systems

Foxconn is also a strategic partner in NVIDIA’s DSX Go-To-Market initiative, integrating digital twins, DSX SimReady assets, simulation validation, and modular data center design to accelerate deployment and optimize efficiency of next-generation AI factories.

Agentic AI, Edge, and Next-Gen Connectivity

This year, Foxconn, together with its partners, is showcasing its latest achievements in the fields of agentic AI, edge computing, and next-generation communications. In collaboration with Intel, exhibits demonstrate the AI Head Node Server, GNR-AP / CWF-AP Server, 1.6T OSFP 2xDR4 high-speed optical communication solution, and next-generation 5G DU Edge Server, further strengthening Foxconn's presence in AI data centers, high-speed networks, and next-generation communications infrastructure.

In addition, Foxconn also showcased its collaboration with AMD, including the Ingrasys x AMD solution, the GB2281A platform, and the AMD Instinct™ MI350P PCIe Card.

Space Data Centers

Together with Ramon.Space, Foxconn is demonstrating related applications and next-generation communication technologies, including solutions such as NuComm, NuPod, NuBox, and edge computing devices. The efforts expand the application of AI infrastructure to fields such as space computing, low-Earth orbit satellite communications, and next-generation data centers.

Optical and Core Components

Also on display were key optical transmission technologies such as CPO (Co-Packaged Optics), 1.6T optical communication modules, CPC Cable to XPO, and ELSFP for CPO Switch Solution, as well as the latest hardware from L1-L12 manufacturing, including key components such as PCBs, water-cooled plates, and branch pipes, highlighting Foxconn’s vertical integration of AI servers and next-generation data center architecture.

AI Compute Applications

As one of the few NVIDIA Cloud Partners (NCP) in Taiwan with AI Factory infrastructure capabilities, the supercomputing center led by Foxconn’s Visionbay.ai integrates strengths in AI server manufacturing, thermal solutions, supply chain, and data centers, offering a GPU cloud platform for the AI agent ecosystem. This enables enterprises and developers to rapidly deploy, manage, and scale AI agent applications.

AI Models and Agent Systems

Running concurrently to COMPUTEX, the NVIDIA GTC Taipei will see Hon Hai Research Institute introduce FoxBrain 2.0, which integrates NVIDIA Nemotron Omni architecture and multimodal training. It evolves from “understanding” to “thinking, planning, and execution,” enabling Physical AI capabilities and extends to Smart Manufacturing, Smart EV, and Smart City, accelerating the implementation of industrial-grade embodied AI.

Foxconn also introduced CoDoClaw, a clinical AI agent system built on NVIDIA NemoClaw. Centered around a “lobster agent,” it integrates multiple AI agent capabilities into medical workflows, enabling multi-agent collaboration in intelligent healthcare automation and serving as a “super clinical assistant” for physicians.

Robotics and Physical AI

Foxconn’s Nurabot has completed field validation and is being deployed across multiple hospitals and long-term care settings. According to actual deployment data, Nurabot can perform 75 to 80 tasks per day and help reduce the workload of nursing staff by about 30% through clinical support work such as drug delivery and specimen transport.

A collaborative surgical robot, that integrates multimodal perception, surgical scene understanding, and task reasoning, is jointly developed with Kawasaki Heavy Industries and partners, utilizing:

• NVIDIA Isaac for Healthcare (Agent-Ready Rheo Blueprint)
• NVIDIA Isaac GR00T Vision-Language-Action (VLA) architecture

On the manufacturing side, Foxconn demonstrated a wheeled humanoid robot solution integrating NVIDIA Isaac Teleop, MOMClaw, a teleoperation framework, and Physical AI technology. Applications include precision industrial assembly tasks such as pick-and-place, dual-arm collaboration, and force-controlled screw fastening. The system integrates vision, motion control, path planning, and force feedback, and supports development environments like Isaac Sim, Isaac Lab, and ROS 2, accelerating deployment and enhancing flexible manufacturing automation.