Taipei: U.S. semiconductor company AMD announced a significant investment exceeding US$10 billion in Taiwan's AI supply chain ecosystem while simultaneously increasing production of its next-generation server processors utilizing Taiwan Semiconductor Manufacturing Co's (TSMC) advanced 2-nanometer process technology.
According to Focus Taiwan, AMD's press statement revealed that the investments aim to expand collaborations with Taiwanese companies specializing in chip packaging, substrates, and server manufacturing, as global demand for AI infrastructure continues to rise. AMD Chair and CEO Lisa Su emphasized the synergy between AMD's high-performance computing technologies and Taiwan's manufacturing ecosystem, combined with global partnerships, to develop integrated, rack-scale AI infrastructure for future AI systems.
A key aspect of the investment strategy involves advanced packaging technology for AI chips, particularly the development of next-generation 2.5D bridge interconnect technology, known as Elevated Fanout Bridge (EFB). AMD is working with Taiwan-based ASE Technology Holding Co. and Siliconware Precision Industries Co. (SPIL) to enhance EFB packaging, aiming to improve chip interconnect bandwidth and power efficiency for future EPYC processors.
Additionally, AMD reported achieving a milestone with Powertech Technology Inc. (PTI) by qualifying the industry's first 2.5D panel-based EFB interconnect. This development is expected to facilitate more scalable and cost-effective AI chip production.
The investments are linked to AMD's forthcoming Helios AI platform, described as a production-ready rack-scale AI infrastructure system set for deployment beginning in the second half of 2026.
In a separate announcement, AMD disclosed that production of its next-generation EPYC server processor, codenamed "Venice," has commenced in Taiwan utilizing TSMC's advanced 2nm process technology, marking it as the industry's first high-performance computing product to do so. The increasing demand for "agentic AI" workloads underscores the growing role of CPUs in managing data movement, networking, storage, and system orchestration within AI data centers.
AMD also plans future production of "Venice" processors at TSMC's Arizona facility, reflecting its strategy to diversify its advanced manufacturing capabilities.