Separator

ASMPT and IBM Enhance Chiplet Bonding for AI Packages

Separator

ASMPTASMPT and IBM have announced a renewed agreement to extend their collaboration in developing cutting-edge chiplet packaging technologies. The partnership aims to advance thermocompression and hybrid bonding technologies for chiplet packages using ASMPT’s next-generation Firebird TCB and Lithobolt hybrid bonding tools.

Chiplets deconstruct system-on-chips (SOCs) into smaller components, which can be packaged together to function as a single system. This approach offers benefits such as improved energy efficiency, faster system development cycles, and reduced costs. However, advancements in packaging are essential to transition chiplets from research to mass production swiftly, driven by the rapid innovation pace in AI computing.

The latest agreement builds on an existing collaboration between ASMPT and IBM. Last year, their joint efforts led to a new hybrid bonding approach that optimized bonding quality between chiplets. This ongoing partnership will continue to develop bonding technologies crucial for chiplet packages.

“IBM has been at the forefront of developing advanced packaging technology for the age of AI”, said Huiming Bu, Vice President of IBM Semiconductors Global R&D and Albany Operations, IBM Research. “We are proud to continue our work with ASMPT to advance chiplet packaging technology to pave the way for smaller, more powerful, and more energy-efficient chips”.

Lim Choon Khoon, Senior Vice President at ASMPT, expressed similar enthusiasm about the collaboration. “We are excited to build on our strong relationship with IBM to drive the frontiers of advanced packaging in tandem with accelerating innovations in artificial intelligence. We are pleased to work with IBM to advance next-generation packaging and heterogeneous integration solutions for the AI era”.

This partnership between ASMPT and IBM marks a significant step forward in the development of advanced chiplet packaging technologies, promising to enhance the efficiency and performance of AI-driven systems.

Current Issue