Steve Hou
Head of Research at Silicon Data, AI Compute Markets
Amkor stands to gain as outsourced packaging and test providers become central to advanced chips, and its 10-year TSMC deal locks that in. Per @wallstengine, TSMC will procure advanced packaging and testing services from Amkor's Arizona campus, linking TSMC's US chip manufacturing with Amkor's US capacity. The deal is good for CoWoS and great for Amkor, but pressures Intel, whose EMIB moat rests partly on on-shore availability to US fabless customers. x.com x.comÂ
Amkor and ASE gain as TSMC's advanced-packaging spillover grows, making this the season for outsourced packaging. The share of spillover customers at TSMC has risen drastically, so the Amkor partnership was the viable option, and ASE has huge opportunities too. A post the author replies to calls packaging the biggest bottleneck, with chiplets, compute-memory stacking and silicon photonics driving growth. x.comÂ
Intel is overvalued after running up about 5x from $20, with more capex, low operating margins and a run-up driven largely by Trump's promotion rather than fundamentals. TSMC earns over 30% operating margins and is ahead of schedule in Arizona, so Intel has no US fab monopoly. Trump leaves office in two years, and these fabs are unlikely to meaningfully improve on existing 2nm yields. It could still be a great trade. x.com x.com x.com x.com x.comÂ
AMD and Nvidia face a growing capital burden from substrates without the benefit of foundry margin stacking. Each is committing $1 billion to the substrate ecosystem; AMD has also publicly committed $10 billion to its Taiwan supply chain, more than three times its Q2 net income. That capital is unavailable for R&D and product development, while their gross margins are flat or shrinking. Intel's gross margin is growing as customers prepay its foundry for substrates. x.comÂ
Intel's foundry may be closer than the market assumes because its UMC 12nm partnership brings external-customer experience and a more customer-friendly design kit. Musk said Terafab plans to use Intel 14A, with Tesla and SpaceX building and running the fabs themselves; that would be an important outside test. Intel still must prove yields, volume and schedules, while Terafab's commercial terms remain unclear. Still bullish, but not enough to add now. x.comÂ
Intel's valuation is at all-time highs while its business condition is near all-time lows. It is more a CPU company than a foundry, and TSMC holds huge scale, capex and process advantages while building substantial US capacity, so the recent rally makes no sense. x.com x.comÂ
TSMC may lose pricing power as performance gains shift from shrinking transistors to stacking, interconnects and optics, while logic folding could make older processes competitive. Despite strong leading-edge demand, investors may overestimate the profit share its node leadership can retain. Its cautious expansion and past reluctance toward advanced packaging suggest a risk-averse culture that protects returns but could leave it late to disruptive technologies and key customers. x.comÂ
TSMC is not a favorite longer term because the US administration and US companies have a stated goal of building a domestic source of silicon supply, and relationships toward that are already being forged. x.comÂ
SK hynix reportedly considering Intel Foundry for HBM4E base dies is a negotiating tactic against TSMC, not a real win. Companies, Nvidia included, keep 'considering' Intel and then go quiet, because they know Intel still falls short. x.comÂ
Architecture changes could lift TSMC wafer demand. Shifting from 12-high to 8-high lifts volumes; an on-device AI phone upgrade cycle would be enormously bullish, while ASIC gains over Nvidia help because ASICs use more wafers. TSMC is arguably more vital to AI buildout than MAG stocks and a reasonable extended-AI-cycle play at a not-crazy multiple, barring big macro events. Margins may not expand as capex rises; a Terrafab partnership, if it happens, would somewhat ease Intel competition risk. … Show more