The global AI surge has cast an unexpected spotlight on Taiwan's hardware industry. On July 1, the Project TAME (TAiwan Mixture of Experts) initiative was launched, aiming to establish Taiwan as a major AI player beyond its role as a supply chain provider.
Led by a consortium including Chang Chun Group, Pegatron, and Unimicron, in collaboration with National Taiwan University's Departments of Computer Science and Information Management and LegalSign, this groundbreaking initiative seeks to enhance Taiwan's AI capabilities. Project TAME will develop the world's first traditional Chinese language expert model and promote open-source development with support from Nvidia Taiwan.
Expertly tailored for Taiwan
Ethan Hsiao, Nvidia Taiwan's Business Director, emphasized the goal of uniting industry and academia to develop multi-modal and multi-language AI applications, fostering a robust AI ecosystem tailored to Taiwan's needs. Lin Shean-Tung, Chairman of Chang Chun Group (Dalian), highlighted the rapid expansion of China's petrochemical industry and the necessity for AI-driven predictive analysis to stay competitive.
Chang Chun Group's AI strategy includes developing a comprehensive command center encompassing production, sales, human resources, development, and finance. The company is also integrating ChatGPT with robotics for general applications, aiming to achieve effective learning through demonstration training.
From automation to autonomous
Pegatron is actively exploring AI applications in manufacturing, transitioning from automation to autonomous operations. The company has implemented PEGA AI and PEGAVERSE for various applications, including site simulation and defect detection, and plans to expand AI integration into R&D stages.
Pegatron expressed that participating in Project TAME aims to accelerate the application of generative AI across industries through innovative collaboration, combining expert data from different sectors with academic expertise in large language models (LLM).
As AI and digital twins become essential for advancing from automation to autonomous operations, Pegatron will continue to invest in related technologies and applications. Project TAME aligns with the sovereign AI trend, recognizing the need for extensive data, models, and computational power, which require significant investment beyond the reach of individual companies.
The initiative aims to unite Taiwan's industry and academia under Nvidia's developer program, developing a traditional Chinese language expert model in an open-source format for collaborative use across sectors. As more companies seek computational power for model training and application, Project TAME aspires to catalyze AI software development in Taiwan, leveraging LLM training across various industries.
