InP substrate demand and pricing still support the AI optics supply chain. Yunnan Germanium reports higher InP volumes from AI data centers and optical communications, plus higher prices. Q3 revenue rose 134.6% year on year, near AXT's guided 136%, but also includes germanium. IQE's H1 photonics grew 45%, partly from US defense funding. AXT's $66m Q3 guide covers only licensed exports or sales needing no license; a miss would more likely reflect licensing. x.com 

IQE's new multi-year InP epiwafer supply deal with Tower Semiconductor shows its product is irreplaceable, adding a third Tier-1 foundry anchor in the AI photonics stack after Lumentum and MACOM. IQE is the non-Chinese InP source Tower needed, a geopolitical moat, and is designed into the optical roadmap from 200G pluggables to 400G-per-lane modulators and optical circuit switches. Tower, which IQE sued for IP theft in 2022, even granted IQE a royalty-free patent license. x.com 

IQE's outsourced epitaxy business is structurally disadvantaged because demanding optics customers want the whole chain under their own control. Big integrated players such as Lumentum and Applied Optoelectronics run their own reactors and want to control manufacturing conditions throughout. Shipping wafers to an outside fab for epi growth and back falls short of that ideal, so regardless of IQE's execution the business model is unappealing, even as @crux_capital_ turns more positive. x.com 

IQE is levered to whichever optical architecture wins: GaAs epiwafers if VCSELs gain ground, which seems likely, and indium phosphide epiwafers via Tower Semiconductor if silicon photonics wins. It built the world's largest VCSEL epiwafer facility in Newport and has automotive VCSEL volume orders plus growing defense demand. On June 15 it signed a multi-year InP supply deal with Tower for 200G/lane and 400G/lane products, settling their IP disputes. A small slice of either market would matter. … Show more

IQE's new multi-year indium phosphide epiwafer supply deal with Tower Semiconductor puts a 5x-10x move from March 2026 levels in the works, and IQE is the pick over Tower itself. Per Wall St Engine, the wafers support Tower's silicon photonics platforms for AI data center optics, including 200G/lane pluggable transceivers, 400G/lane modulator prototypes and optical circuit switches. The two also settled their IP dispute, with Tower granting IQE a worldwide royalty-free license. x.com  x.com 

IQE is hard to call undervalued because its price already embeds a mature-margin scenario two-plus years out. At a 15% margin and 20% yearly growth, FY28 EBITDA is about GBP25 million, worth about GBP500 million at a generous 20x, versus a GBP520 million market cap at 40p. IQE earns about 3% adjusted EBITDA margins versus LandMark's high-margin epi, MACOM's convertible and warrants carry about 76 million shares of overhang, and InP supply constraints threaten its H2-weighted guide. x.com  x.com  … Show more

IQE is reaching its financial inflection point as InP demand for AI and data centers drives growth. It raised 2026 revenue growth guidance from 20% to over 30% after H1 revenue of £64m versus £45m, and forecasts low-teens £m adjusted EBITDA. MACOM's investment left it debt-free with £42m of cash. The 30% target looks deliberately conservative after past over-optimism; as @mkfilko notes, it implies H2 revenue below H1, which seems unlikely. x.com  x.com 

MACOM is a hold amid the AI sector sell-off: its investment in IQE de-risks operations a lot, but Coherent and Lumentum are better ways to play the 800G and 1.6T transition. x.com 

IQE is held above its fundamental fair value by optimistic expectations, and its takeover thesis is effectively dead. With MACOM as anchor investor holding board seats and exclusive long-term supply deals, a bidding war is practically ruled out. The 19.8p placing, a 58% discount, takes dilution since early 2025 to about 42%, and the £81 million mostly repays debt, so capex needs another raise within 18-24 months. If momentum fades the stock will fall; if it strengthens, it can keep rising. x.com 

IQE is priced for skepticism it no longer deserves, while Sivers Semiconductors is priced for perfection it has not delivered, though both sit in the same AI optical chain. Sivers trades at about 85x sales with negative EBITDA, and Q1 revenue missed. IQE trades near 4-5x sales, is EBITDA positive, guides 20%+ growth to at least £117M in 2026, has cleared bank debt with MACOM's backing, and trades far below peer LandMark at about 70x forward. Sivers is worth revisiting only if its price resets. … Show more