TSMC’s First-Generation CoPoS May Not Start With Glass

Executive Summary

A widely held assumption in Asia’s advanced packaging supply chain is that TSMC’s upcoming Chip-on-Panel-on-Substrate, or CoPoS, platform will help trigger large-scale adoption of glass-core substrates. According to a July 6 report by CommonWealth Magazine, that assumption may be early.

The report says industry insiders indicate TSMC’s first-generation CoPoS, expected in the 2028–2029 timeframe, may launch without glass substrates. If accurate, that would not undermine the long-term case for glass in advanced packaging, but it would change the timing of the transition that many supply-chain participants and investors appear to have been anticipating.

For Taiwan and Japan, the issue is strategically important because CoPoS sits at the intersection of AI infrastructure demand, advanced substrate technology, and packaging capacity planning. A glass-free first generation would suggest that the market may be pricing in a materials shift faster than TSMC’s near-term roadmap supports.

Watch the Short Brief

This short explains why a potentially glass-free first generation of TSMC CoPoS would change timing expectations across Asia’s advanced packaging supply chain.

Key Developments

According to the available source information, the central new development is not the existence of CoPoS itself, but the reported possibility that its first generation may be glass-free. CommonWealth Magazine reported that industry insiders expect first-generation CoPoS in 2028–2029 and said it may not initially use glass substrates.

That matters because CoPoS is broadly viewed as an important step in advanced packaging for larger and more demanding AI and high-performance computing devices. CoWoS has already become a major part of the AI packaging discussion, and CoPoS is often seen as part of the next phase of scaling package size and integration through panel-based processes.

The report also places the implications in a Taiwan-Japan supply-chain context. Companies named in that discussion include TSMC, Innolux, Ibiden, and Gudeng Precision. Based on the available information, the most defensible conclusion is that these companies are relevant to the broader packaging and substrate ecosystem around CoPoS, rather than that any specific revenue or investment outcome has been confirmed.

Strategic Analysis

The important point is timing. The long-term technical rationale for glass-core substrates remains part of the broader industry debate around next-generation packaging. But a technology can be strategically important without being adopted in the first commercial cycle.

If the CommonWealth report proves accurate, the immediate implication is that the market’s “glass revolution” narrative may be ahead of TSMC’s first deployable CoPoS configuration. That would matter for capital allocation across Asia’s semiconductor supply chain, particularly in Taiwan and Japan, where packaging, substrate, and related equipment ecosystems are deeply tied to leading-edge logic demand.

One strategic implication is that investors and operators may need to separate two questions that are often blended together: whether glass becomes important in advanced packaging over time, and whether TSMC’s first-generation CoPoS is the event that brings that shift into volume production. Those are not necessarily the same thing.

This distinction could reshape how the market interprets near-term winners. A delay in glass adoption at the first-generation CoPoS stage would not necessarily mean weaker CoPoS momentum overall. Instead, it may indicate that packaging scale-up and substrate transition are moving on different schedules. That creates a more nuanced outlook for suppliers tied to panel processes, substrates, materials, and precision handling.

For Asia’s technology landscape, the broader lesson is that advanced packaging roadmaps are increasingly becoming as important as front-end process leadership in determining supply-chain positioning. In AI semiconductors, packaging choices influence not just performance potential, but also manufacturing readiness, qualification risk, and the timing of ecosystem returns on investment.

Investor Takeaway

The reported possibility of a glass-free first-generation CoPoS introduces a timing risk rather than a definitive change in the long-term direction of advanced packaging.

For investors, the key question is whether current expectations for glass-core substrate adoption are anchored to a transition that may arrive later than assumed. That makes roadmap verification more important than narrative momentum. Readers should watch for future disclosures or reporting that clarify CoPoS substrate choices, as well as any change in how Taiwan and Japan supply-chain companies discuss packaging-related capex, development priorities, or commercialization timelines.

The main strategic conclusion is straightforward: even if glass remains part of the industry’s future, the first commercial phase of CoPoS may not be the catalyst the market has been expecting. If that view is confirmed, some of the most immediate consequences would likely be felt not in end-demand for AI compute, but in the timing of returns across Asia’s advanced packaging supply chain.