After a 63% gain in the past three months, Intel has sharply outperformed other major chip names (AMD +30%, TSMC +22%, Nvidia -2.5%). The article frames Intel’s recent rerating as driven not by near-term earnings but by its potential role as a U.S.-based leading-edge foundry and an alternative supplier on the advanced packaging layer. Advanced process nodes and packaging are now described as key bottlenecks in Artificial Intelligence chip production, prompting hyperscalers and institutional investors to seek technological and geographic diversification beyond traditional chip-level diversification.
Multiple external customer orders are cited as a tailwind. Tom’s Hardware, citing analyst Ming-Chi Kuo, reports that Apple has signed an NDA with Intel and received the 18A PDK 0.9.1GA, with potential volume shipments as early as mid-to-late 2027. On the packaging side, Intel’s EMIB is gaining traction; reports suggest Google plans to use EMIB in its 2027 TPU v9 and market rumors point to Meta’s MTIA AI accelerators adopting Intel’s advanced packaging. TrendForce data in the article quantify package scaling: CoWoS-S is estimated to support reticle-size expansion of up to about 3.3×, CoWoS-L around 3.5× today and potentially up to 9× by 2027; EMIB-M already supports about 6×, and is expected to reach 8-12× in the 2026-2027 time frame. By eliminating an intermediate layer, EMIB is presented as a more cost-effective option for ultra-large Artificial Intelligence packages compared with CoWoS-only solutions.
Despite the strategic narrative, the piece emphasizes short-term earnings pressure and execution risk. Most marquee orders only start to matter around 2027, so investors should not expect significant foundry-driven earnings uplift in 2025-2026. Intel’s 3Q results showed the foundry segment lost roughly US?.3 billion, with new nodes such as 18A still in ramp-up. The client computing business operates in a saturated market, while data center and Artificial Intelligence segments face intense competition and primarily serve as cash-flow providers today. Market consensus cited in the article expects 2025 full-year adjusted EPS of about ?.34 and 2026 of about ?.63; at the current share price that implies ~100x 2025E P/E and 60-70x for 2026E.
Looking beyond 2027, the article frames the outlook as binary. If Intel can scale 18A into high-volume, competitive production and move the foundry business from deep losses toward breakeven and profitability, the current valuation could be justified as payment for a structurally valuable U.S. strategic asset. If external orders ramp more slowly and foundry losses persist, the foundry push will act as a drag on group valuation. The piece positions EMIB and packaging diversification as the immediate technical lever that could reconfigure relationships in the foundry ecosystem and chip supply chains.
