Smartkem (Nasdaq: SMTK), a developer of semiconductor materials, has entered into a preliminary Joint Development Agreement (JDA) with Manz Asia to co-develop advanced dielectric ink solutions for next-generation chip packaging, with a particular focus on artificial intelligence (AI) and high-performance computing applications.
The agreement, while non-binding, marks a strategic step toward addressing key bottlenecks in AI chip packaging. The collaboration aims to leverage Smartkem’s expertise in organic semiconductor materials and Manz Asia’s inkjet-based packaging equipment to deliver scalable solutions for both wafer-level and panel-level packaging technologies.
“This joint development agreement builds on our recent demonstration at SEMICON SEA 2025,” said Ian Jenks, Chairman and CEO of Smartkem. “We aim to co-develop high-performance, scalable packaging solutions that lower the cost and increase yield for AI accelerators in data centers. Our materials, paired with Manz’s precision inkjet technology, can help unlock new efficiencies at both the wafer and panel levels.”
The collaboration seeks to address growing demand for 12-inch wafer-level packaging and explore panel-level techniques that could extend chip integration beyond conventional wafer constraints. This evolution is seen as critical to meeting performance and energy demands in data centers where AI accelerator deployment is expanding rapidly.
Robert Lin, General Manager of Manz Asia, highlighted the strategic potential: “Smartkem’s dielectric ink formulations, which offer high-resolution patterning and chemical durability, are well-positioned to improve top metal insulation and redistribution layer (RDL) applications. This partnership could accelerate the industrial adoption of advanced inkjet processes for semiconductor back-end-of-line (BEOL) manufacturing.”
Advanced dielectric inks are crucial in AI chip packaging, particularly in managing insulation and interconnect complexity. Panel-level packaging, a less explored frontier, could allow for increased chip density, faster interconnects, and improved thermal management—key to scaling AI infrastructure sustainably.
While the JDA is preliminary and non-binding, both companies signal intent to finalize terms and advance development, recognizing the potential to reshape the economics and scalability of AI hardware manufacturing.
