Foxconn's FoxBrain AI to transform manufacturing and supply chain operations

Powered by 120 Nvidia H100 GPUs, FoxBrain, built on Meta's Llama 3.1 architecture, is Taiwan’s first AI model optimized for traditional Chinese, offering advanced reasoning for Foxconn's AI-driven automation

Taiwanese electronics giant and key Apple supplier Foxconn has officially launched its inaugural large language model (LLM), FoxBrain, marking a significant step in the company’s push toward smart manufacturing and artificial intelligence (AI) integration within its operations. This FoxBrain launch demonstrates Foxconn’s commitment to driving innovation through advanced technology to enhance manufacturing processes and streamline supply chain management.

Developed with the power of 120 Nvidia H100 GPUs, FoxBrain completed its training in just four weeks. Built on Meta’s Llama 3.1 architecture, this Foxconn AI model is the first of its kind in Taiwan, specifically optimized for traditional Chinese and Taiwanese language styles. The model also features advanced reasoning capabilities, positioning it as a key asset for Foxconn’s AI-driven workflow automation.

FoxBrain to drive AI-powered collaboration

Foxconn’s internal testing has shown that, while there is a slight performance gap compared to China's DeepSeek distillation model, FoxBrain’s overall performance aligns closely with leading global LLM standards. Initially, FoxBrain will be used for internal applications such as data analysis, decision support, document collaboration, mathematical reasoning, problem-solving, and code generation, all aimed at advancing Foxconn smart manufacturing.

Foxconn plans to extend the FoxBrain technology by collaborating with technology partners to promote AI adoption across manufacturing, supply chain management, and intelligent decision-making processes. This move will drive further AI integration, contributing to Foxconn’s larger vision of revolutionizing the industry through Foxconn innovation.

Nvidia has played a crucial role in the development of FoxBrain, offering substantial support through its Taiwan-based supercomputer, Taipei-1, and providing technical consulting during the model’s training phase. Taipei-1, Taiwan's largest supercomputer, is owned and operated by Nvidia in Kaohsiung.

Further details about FoxBrain’s capabilities will be unveiled during Nvidia’s upcoming GTC developer conference in mid-March.