Hon Hai Technology Group, better known as Foxconn, this week said they plan to jointly build an advanced computing center in Kaohsiung, Taiwan. At the core of the center will be the NVIDIA Blackwell platform. The companies made the announcement at Computex 2024. 

NVIDIA said the cutting-edge computing center will be anchored by the super chip GB200 servers and consist of a total of 64 racks and 4,608 GPUs. The electronics manufacturer will contribute its production scale and said it expects to complete the center by 2026. 

The companies said their latest collaboration demonstrates their commitment to building servers to drive artificial intelligence, electric vehicles (EVs), smart factories, smart cities, robotics, and more. 

“A new era of computing has dawned, fueled by surging global demand for generative AI data centers,” stated Jensen Huang, founder and CEO of NVIDIA. “Foxconn stands at the forefront as a leading supplier of NVIDIA computing and a trailblazer in the application of generative AI in manufacturing and robotics.”

“Leveraging NVIDIA Omniverse and Isaac robotics platforms, Foxconn is harnessing cutting-edge AI and digital twin technologies to construct their advanced computing center in Kaohsiung,” he added.

Cooperation continues for Foxconn with new superchip 

This isn’t the first time Foxconn and NVIDIA have collaborated. The company has worked closely with NVIDIA on various product development projects. NVIDIA said Foxconn has excellent vertical integration capabilities and is a vital partner for the new GB200 Grace Blackwell Superchip. 

The superchip connects two NVIDIA B200 Tensor Core GPUs to the NVIDIA Grace CPU over a 900GB/s ultra-low-power NVLink chip-to-chip interconnect. The company said that for the highest AI performance, GB200-powered systems can be connected with the NVIDIA Quantum-X800 InfiniBand and Spectrum-X800 Ethernet platforms.

---

Now accepting session submissions!

---

NVIDIA and Foxconn plan for the facility

The partners said that NVIDIA’s AI technology will drive Foxconn’s three smart platforms: Smart Manufacturing, Smart EV, and Smart City. The new facility will use NVIDIA Omniverse to create digital twins for these platforms.

Foxconn plans to use image-recognition technology combined with its autonomous mobile robots (AMRs) to provide optimal capacity utilization in smart manufacturing. The companies said they will also take on production-line planning, which will encompass the existing manufacturing of AI servers and electric vehicle assembly plants. 

Foxconn subsidiary Foxtron’s Qiaotou automotive manufacturing facility will be one of Foxconn’s benchmark AI factories. Currently under construction, the site will use digital twins connected to cloud technologies. The company also hopes to collaborate between virtual and physical production lines.

In addition, the facility is set up with digital real-time monitoring to ensure high-quality manufacturing of an electric bus. 

NVIDIA and Foxconn plan to collaborate on future electric vehicle models designed by Foxconn. Currently, the company is negotiating projects with traditional European and American automakers. The partners also said they plan to develop a “cabin-driving-in-one” smart travel system.