Korea invented panel-level packaging but may lose the race to Taiwan — and the reason is surprisingly mundane: no one can agree on the size of a rectangle.
PLP swaps round wafers for flat panels. Better yields, better thermals, perfect for massive AI chips. Samsung shipped it first for phones and wearables. Now they're scaling into AI and HPC.
But there's no standard. Over 30 different panel sizes floating around — 310×310 mm, 415×510 mm, 600×600 mm, 700×700 mm. Some lines converted from substrates, others from LCD fabs. Equipment makers and materials suppliers have to support all of them. That kills economies of scale, slows R&D, tanks yields.
Meanwhile TSMC picked one size — 310×310 mm — for its pilot line. The entire supply chain is rallying behind it. OSATs, equipment vendors, materials firms. They're leveraging 300 mm wafer knowledge and moving fast. Korea has no equivalent anchor.
Some in the industry say Korea should just adopt TSMC's standard, lower friction, win early customers. But that's the trap. Follow now, and you cement yourself as a follower forever. Materials suppliers get absorbed into the TSMC ecosystem. Taiwan owns the roadmap. Korea becomes a satellite.
The smarter play: Samsung locks in stable high-end yields first, lands a few marquee AI orders, then sets its own clear standard with a multi-year roadmap. Give domestic suppliers volume visibility and co-investment certainty. Build the ecosystem around Korean leadership, not Taiwanese momentum.
This isn't about technology. It's about who writes the rules. Standardization is boring until it decides who controls the next decade of advanced packaging. Right now, Korea is running out of time to make that call.
PLP swaps round wafers for flat panels. Better yields, better thermals, perfect for massive AI chips. Samsung shipped it first for phones and wearables. Now they're scaling into AI and HPC.
But there's no standard. Over 30 different panel sizes floating around — 310×310 mm, 415×510 mm, 600×600 mm, 700×700 mm. Some lines converted from substrates, others from LCD fabs. Equipment makers and materials suppliers have to support all of them. That kills economies of scale, slows R&D, tanks yields.
Meanwhile TSMC picked one size — 310×310 mm — for its pilot line. The entire supply chain is rallying behind it. OSATs, equipment vendors, materials firms. They're leveraging 300 mm wafer knowledge and moving fast. Korea has no equivalent anchor.
Some in the industry say Korea should just adopt TSMC's standard, lower friction, win early customers. But that's the trap. Follow now, and you cement yourself as a follower forever. Materials suppliers get absorbed into the TSMC ecosystem. Taiwan owns the roadmap. Korea becomes a satellite.
The smarter play: Samsung locks in stable high-end yields first, lands a few marquee AI orders, then sets its own clear standard with a multi-year roadmap. Give domestic suppliers volume visibility and co-investment certainty. Build the ecosystem around Korean leadership, not Taiwanese momentum.
This isn't about technology. It's about who writes the rules. Standardization is boring until it decides who controls the next decade of advanced packaging. Right now, Korea is running out of time to make that call.