AMD announced more than $10 billion in investments across the Taiwan ecosystem to expand AI infrastructure, partnering with ASE, SPIL, and other firms on advanced packaging and manufacturing. The move builds on AMD's long-standing leadership in chiplet architectures and 3D hybrid bonding, aiming to deliver faster AI system deployment at scale.

The investment covers next-generation wafer-based 2.5D bridge interconnect technology, developed with ASE and SPIL to support Venice CPUs. AMD also qualified the industry's first 2.5D panel-based EFB interconnect through PTI, enabling higher bandwidth and better economics for AI deployments.

These packaging advances feed into the AMD Helios rack-scale platform, slated for deployment in the second half of 2026. Sanmina, Wiwynn, and Inventec are building Helios systems with Instinct MI450X GPUs and 6th Gen EPYC CPUs, designed to run larger AI workloads faster while optimizing power efficiency.

The announcement signals AMD's aggressive push to compete in AI infrastructure. Helios targets hyperscaler demands for higher performance-per-watt, and the Taiwan partnerships represent a concrete step to accelerate production-ready AI systems.