TSMC Welcomes Advanced Packaging Competition, Easing Pressure on MediaTek
Alina Collins
TSMC Chairman C.C. Wei told an earnings call that more advanced-packaging competitors are welcome because back-end capacity constraints are capping customer growth — a statement the industry reads as easing pressure on MediaTek over its use of Intel packaging.
Why is TSMC inviting packaging rivals in?
C.C. Wei said on July 16 that tight advanced-packaging capacity — the high-end process of assembling multiple chips into one package — is already capping customer growth.
This means → customers want to buy more wafers, but packaging can't keep up, so TSMC's own foundry revenue hits a ceiling too.
In plain terms = TSMC can't do all the packaging work itself; letting others pick up the slack lets customers grow — and growing customers order more wafers from TSMC.
Why was MediaTek so nervous before?
MediaTek adopted Intel's EMIB packaging — Embedded Multi-die Interconnect Bridge, a technique that buries small silicon bridges inside the substrate to link multiple chips — but was extremely cautious in public, repeatedly stressing that TSMC remains its top partner.
This means → MediaTek's real worry was not the technical choice itself but the risk that TSMC might question its loyalty.
With advanced-node capacity already stretched thin, a downgrade in TSMC priority would hurt MediaTek far more than any packaging diversification could help.
What does Wei's statement change for MediaTek?
The industry reads Wei's remarks as a pressure release for MediaTek — TSMC is not seeking absolute dominance in advanced packaging.
This means → MediaTek can now discuss packaging partners more openly, without fearing it will anger TSMC.
Where is TSMC's real moat?
Industry veterans note that advanced packaging has always been more of an add-on service for TSMC, not its core profit engine — its capex allocation shows packaging expansion running well behind front-end wafer-fab buildouts.
Wei emphasized that TSMC's foundry technology lead cannot be replicated by Intel with government support, nor caught by Samsung with capital spending alone — years of accumulated know-how are the core.
This reflects that TSMC's true priority is foundry-level irreplaceability; ceding some packaging ground does not shake the foundation.
What to watch next?
In the near term, the chance of MediaTek shifting advanced-node orders to another foundry remains very low — TSMC is still the first choice.
In plain terms = MediaTek can shop around for packaging, but for chip manufacturing it cannot walk away from TSMC yet.
The key question going forward: can packaging-level diversification keep advancing without disturbing the foundry-level partnership?
Content is for reference only, not financial advice.