AMD has committed more than $10 billion to strengthen AI infrastructure through strategic partnerships in Taiwan, aiming to enhance advanced packaging manufacturing for next-generation systems. This investment seeks to meet the growing demand for AI infrastructure by advancing technologies that improve performance, efficiency, and deployment speed.
The company's efforts to enhance silicon, packaging, and manufacturing technologies align with its collaborations with major Taiwanese firms such as ASE, SPIL, and Sanmina. Through these partnerships, AMD intends to refine chiplet architectures, integrate high bandwidth memory, and develop 3D hybrid bonding, alongside designing rack-scale systems capable of managing increasingly complex AI workloads.
A key aspect of AMD's investment is the development of EFB-based 2.5D packaging technology, co-created with ASE and SPIL. This architecture aims to increase interconnect bandwidth and improve power efficiency for the 6th Generation AMD EPYC CPUs, codenamed "Venice." These enhancements are expected to enable faster and more efficient AI system deployments while addressing real-world power and cooling constraints.
In another significant advancement, AMD has qualified what it claims to be the industry's first 2.5D panel-based EFB interconnect technology in collaboration with PTI. This innovation is likely to provide high-bandwidth interconnects at scale, greatly enhancing the economic feasibility of deploying AI systems.
The AMD Helios rack-scale platform, set for deployment in the latter half of 2026, will incorporate these advancements. Equipped with AMD Instinct MI450X GPUs, the latest EPYC CPUs, and an advanced networking framework, the Helios platform will receive support from ODM partners like Sanmina, Wiwynn, Wistron, and Inventec. AMD asserts that this platform will boost compute power, interconnect bandwidth, memory capacity, and overall system integration to handle demanding AI workloads.
Dr. Lisa Su, AMD's Chair and CEO, highlighted the urgency of this initiative, stating, "As AI adoption accelerates, our global customers are rapidly scaling AI infrastructure to meet growing compute demand. By combining AMD leadership in high-performance computing with the Taiwan ecosystem and our strategic global partners, we are enabling integrated, rack-scale AI infrastructure that helps customers accelerate deployment of next-generation AI systems."
Sanmina’s Chairman and CEO, Jure Sola, expressed pride in their partnership with AMD, emphasizing their joint commitment to delivering reliable, high-performance solutions. Other partners, including Inventec and SPIL, shared similar sentiments, reinforcing the collaborative effort to provide stable AI and data center systems that address the industry's evolving complexities.
The alignment between AMD and Taiwan’s manufacturing capabilities is poised to reshape AI infrastructure development. With a focus on advanced packaging solutions and innovative technologies, this multi-billion-dollar investment targets immediate demands while establishing a foundation for the future of AI deployment globally. As artificial intelligence evolves, AMD's strategic initiatives signal a significant shift toward integrated solutions that can support the next wave of technological advancements.
The stories that move AI & crypto markets — before the market reacts.
Free. 7am ET. Five stories. 62,400 readers.
