ALMU research

Position in the Physical Layer of AI thesis

ALMU is the newest anchor (promoted from Pivot to Anchor on 2026-05-05 in the manifesto v2 lock). It anchors two adjacent layers: Photonics and Test & Materials. Aeluma builds indium phosphide (InP) quantum-dot photonic integrated circuits. The InP wafer step is the gating constraint on the next generation of optical interconnects (the 1.6T per-lane and 3.2T datacenter optics that hyperscalers need by 2027). Aeluma is one of three companies that can make the substrate at scale.

Quinn's book opened ALMU on 2026-04-22. The 2026-05-05 snapshot had ALMU at +43.97%, with a +19.23% move on May 5 alone after the NASA award.

Recent catalysts (last 30 to 60 days)

NASA award (April 21). $1.4M Phase II SBIR for integrated quantum-dot lasers. The award validates the technology at the federal-research bar and unlocks the next tranche of contracted R&D.
6 government R&D contracts active. DARPA, NRO, AFRL, plus three commercial confidential. The mix is unusual: most photonics startups have one or two federal customers; ALMU has six.
Indium phosphide capacity ramp. Production substrate shipped to the first lead commercial customer in late April. The customer is not yet named but the disclosure language suggests a Tier 1 (Coherent, Lumentum, or POET).
Sub-$50M float discipline. Management has not done a secondary or registered direct since the IPO. Capital efficiency is the antithesis of typical small-cap photonics, which dilutes every 2 years.

The thesis (what has to be true)

1. The next datacenter optics generation (1.6T and beyond) requires indium phosphide PICs at volume by 2027. The 800G generation is silicon photonics; the 1.6T generation is InP. The roadmap is set by Coherent and Lumentum analyst day slides. 2. Aeluma's quantum-dot process scales to production wafers. Today the demo wafers are in pilot. The Phase II NASA work funds the production-scale move. 3. The commercial laser product ships by FY27 (calendar 2027). If the product slips beyond that, the Tier 1 customer designs ALMU out of the next roadmap. 4. The six government contracts continue to fund the R&D side while the commercial side ramps. The federal portfolio is the bridge to commercial revenue, not the destination.

Kill vectors (what would break the thesis)

Commercial laser slip past FY27. This is the load-bearing kill vector. If the lead commercial customer cancels because the product is late, the asymmetric pay disappears and the company falls back to a federal-only revenue model worth a fraction of the current cap.
InP wafer yield problem. Quantum-dot integration at scale is unproven. A yield issue that adds two quarters to production qualification breaks the customer roadmap.
Coherent or Lumentum acquires a competitor. If a larger photonics player buys Aeluma's closest competitor and ramps an alternative InP path faster, ALMU's window closes.
Capital event. Management has been disciplined, but the cash runway needs the next 18 months to land the commercial ramp. A surprise secondary at a low price would mark a discipline break.
Mechanical exit at break. Discipline rule fires on close below $13 for two consecutive sessions.

Layer context

In the 8-layer Physical Layer of AI map, ALMU sits at the intersection of Photonics and the staged v1.5 Test & Materials layer. Sister names in Photonics: SIVEF (SiTime, MEMS oscillators), LITE (Lumentum, datacenter optics, larger and more diversified), LWLG (Lightwave Logic, smaller and earlier). Sister names in Test & Materials: AEHR (Aehr Test Systems), AXTI (compound semis).

ALMU is the highest-conviction smaller-cap anchor in the book. The asym score is high because the cap is small and the customer set is concrete; if the commercial ramp lands, the position re-rates 3-5x. If it slips, the position halves.

Position discipline (the rules at entry)

Trim trigger: +100 percent from cost. Lower threshold than the +200% standard because ALMU is the smaller-cap, higher-volatility anchor; lock gains earlier.
Kill vector exit: break of $13 for 2 consecutive days. Currently at ~$27 (per 5/5 close), so there is meaningful buffer.
4-quarter thesis-fail rule applies. If the commercial laser slips past FY27 guidance or NASA Phase II funding does not advance to Phase III, the 4-quarter clock starts.

Moat 9. Asym 8. Moat is high because the InP quantum-dot process is patented and the federal-research relationships are sticky. Asym is high because the position is small-cap with a concrete near-term ramp.

Real money. Real position. Real receipts.