![Foxconn]()
Foxconn has launched its first large language model, named "FoxBrain," aimed at improving its manufacturing and supply chain management. The company, known for its work with Apple and Nvidia, is leveraging this AI technology to enhance internal operations and streamline decision-making processes.
Development of FoxBrain Using Advanced Technology
The FoxBrain model was developed with the help of 120 Nvidia H100 GPUs and completed in approximately four weeks. It is based on Meta's Llama 3.1 architecture and tailored to work with traditional Chinese and Taiwanese language styles. This marks a significant achievement as Taiwan’s first AI model with reasoning capabilities, making it stand out in the region's AI development.
Performance Highlights of FoxBrain
While Foxconn acknowledged a slight performance gap compared to some models from China’s DeepSeek, it emphasized that FoxBrain’s performance is very close to global standards. The model offers a range of capabilities, including data analysis, decision support, document collaboration, problem-solving, and mathematical computations, allowing it to serve multiple functions within the company.
Expansion Plans for AI Technology in the Future
Foxconn's immediate plans for FoxBrain are focused on internal applications, but the company also has plans to expand its use in the future. By collaborating with technology partners, Foxconn hopes to broaden the model’s scope to include more areas such as intelligent decision-making and AI integration in manufacturing and supply chain management.
Nvidia’s Support and Upcoming Announcements
Nvidia played a key role in FoxBrain's development by providing technical support and access to Taiwan's largest supercomputer, “Taipei-1.” Foxconn is set to share more details about the model during Nvidia’s upcoming GTC developer conference in mid-March, where it will discuss the AI model's progress and future applications.