Glass substrate supply chain splits into separate timelines

Glass substrate activity has shifted from evaluation to qualification as Absolics runs AMD qualification at its Georgia fab, Samsung Electro-Mechanics reportedly ships samples tied to Apple’s supply chain, and TSMC works to complete a next-generation panel packaging pilot line within the year. The investment question is less about whether glass has promise and more about where it sits in the package, because demand reaches materials, equipment, and substrate makers on different schedules.

AI accelerator packaging has exposed limits in organic substrates, especially warpage and area. Glass can reduce warpage by tuning its coefficient of thermal expansion closer to silicon, while panel-level packaging can improve large square package manufacturing compared with round wafer-level processing. TSMC’s CoPoS is framed as part of that panel transition, with the most firmly identified glass role currently on the temporary carrier side.

The phrase “glass substrate” covers four separate categories: glass core substrate, glass interposer, temporary carrier, and glass fiber cloth. Glass core substrates use through-glass vias and remain technically difficult because drilling, copper filling, and metallization can damage brittle glass. Glass interposers are a separate, harder track. Temporary carriers could see the earliest demand as PLP expands, while T-glass benefits current organic substrates and follows a different cycle from glass core adoption, which is described as 2027 at the earliest.