Intel and Amkor team up to scale emib packaging production

Intel has started emib assembly at Amkor's Incheon facility to meet surging orders from large Artificial Intelligence customers, outsourcing work to boost output while keeping some capacity for its own products.

Intel has begun emib assembly at Amkor’s facility in Incheon, South Korea as part of a move to scale packaging output quickly in response to unexpected market demand. The company is working with long-time osat partner Amkor to accelerate production volumes and avoid the delays that internal expansion would bring. According to the report, some emib capacity will be held for Intel’s own upcoming products while the rest helps meet orders from large Artificial Intelligence customers.

The arrangement underscores the current capacity crunch for advanced heterogeneous packaging, with hyperscalers and Artificial Intelligence chip companies placing large orders for cowos and similar services. While TSMC has been a primary choice for many high-density assemblies, growing interest in Intel’s emib and foveros options is prompting partners such as MediaTek, Google, Qualcomm, and Tesla to consider alternatives. Outsourcing to Amkor reduces lead times compared with shifting production solely within the United States and positions emib as a potential external revenue source ahead of Intel’s next process node launches.

This collaboration builds on the Intel-Amkor partnership announced in late May of this year and represents a tactical step to capture near-term demand for advanced packaging. By leveraging Amkor’s capacity, Intel can respond quickly to surges from major customers while retaining the flexibility to allocate capacity to its own roadmap. The move illustrates how packaging partnerships are being used to bridge immediate supply gaps as the industry adapts to increasing demand for complex assemblies.

55

Impact Score

Micron reportedly drafting new NULL.6 billion HBM plant in Hiroshima

Micron is reportedly planning a major expansion at its Hiroshima site tied to a NULL.6 billion investment and a new high-bandwidth memory plant. Sources say the move is driven by rising demand for Artificial Intelligence-oriented memory and possible subsidies from METI totaling 500 billion yen.

TSMC showcases custom C-HBM4E as N3P logic dies target double efficiency

TSMC detailed moves for its HBM4 generation, including a shift of custom C-HBM4E logic dies to N3P and base dies to N12 to cut operating voltages and boost efficiency. The company also outlined packaging roadmaps such as CoWoS-L to support up to 12 HBM stacks for 2026 Artificial Intelligence parts and confirmed customers including Micron and SK Hynix.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.