×
Foxconn to build Taiwan’s fastest AI supercomputer with Nvidia
Written by
Published on
Join our daily newsletter for breaking news, product launches and deals, research breakdowns, and other industry-leading AI coverage
Join Now

Foxconn and Nvidia collaborate on Taiwan’s fastest AI supercomputer: Foxconn, in partnership with Nvidia, is set to construct Taiwan’s most powerful supercomputer using Nvidia’s cutting-edge Blackwell GPU architecture.

Project specifications and performance:

  • The supercomputer, named Hon Hai Kaohsiung Super Computing Center, was announced at Hon Hai Tech Day.
  • It will feature the GB200 NVL72 platform, comprising 64 racks and 4,608 Tensor Core GPUs.
  • The system is expected to deliver over 90 exaflops of AI performance, making it the fastest in Taiwan.
  • Construction has begun in Kaohsiung, Taiwan, with the first phase expected to be operational by mid-2025 and full deployment targeted for 2026.

Intended applications and strategic importance:

  • Foxconn plans to use the supercomputer for breakthroughs in cancer research, large language model development, and smart city innovations.
  • The project aligns with Foxconn’s “three-platform strategy” focusing on smart manufacturing, smart cities, and electric vehicles.
  • The supercomputer will support Foxconn’s efforts in digital twins, robotic automation, and smart urban infrastructure.
  • It aims to bring AI-assisted services to urban areas like Kaohsiung, positioning Taiwan as a global leader in AI-driven industries.

Technical details and architecture:

  • The GB200 NVL72 platform is optimized for AI and accelerated computing.
  • Each rack features 36 Nvidia Grace CPUs and 72 Nvidia Blackwell GPUs connected via Nvidia’s NVLink technology.
  • The system delivers 130TB/s of bandwidth and functions as a single, unified GPU through Nvidia NVLink Switch.
  • This architecture makes it ideal for training large AI models and executing complex inference tasks on trillion-parameter models in real-time.

Foxconn’s AI initiatives and collaborations:

  • Foxconn is building digital twins of its factories using Nvidia Omniverse.
  • The company was among the first to use Nvidia NIM microservices for developing domain-specific large language models (LLMs).
  • These LLMs are embedded into various internal systems and processes in Foxconn’s AI factories for smart manufacturing, smart electric vehicles, and smart cities.

Industry impact and future outlook: The collaboration between Foxconn and Nvidia represents a significant advancement in AI computing capabilities, with potential far-reaching implications for various industries and Taiwan’s position in the global AI landscape.

  • The supercomputer’s exceptional performance could accelerate breakthroughs in critical areas such as healthcare, urban development, and manufacturing efficiency.
  • This project may inspire similar initiatives worldwide, potentially sparking a new era of AI-driven innovation and competition among tech giants and nations.
  • As AI continues to evolve, the demand for such powerful computing resources is likely to grow, potentially reshaping the landscape of high-performance computing and AI research.
Foxconn to build Taiwan’s fastest AI supercomputer with Nvidia Blackwell

Recent News

Salesforce launches Agentforce 3 with real-time monitoring tools

Enterprise adoption surged 233% in six months as companies move beyond pilot programs.