POET Technologies

FY 2025 net loss ~$63M (Q4 2025 alone -$42.7M); FY 2025 R&D ~$15.9M; aggregate raises Oct 2025 + Q4 2025 + Jan 2026 = ~$525M gross (the largest concentrated capital cycle in POET's history). Ending-2025 cash $430M + January 2026 raise $150M = ~$580M starting-2026 liquidity. At $170M annualized burn run-rate (Q4 2025 implied), runway to ~Q3 2027 absent operating-cash-flow inflection. Bull-pillar reading: management called this a 'decisive transition from development to execution'; bear-pillar reading: same dilution pattern visible 2018-2024 cycles, just larger ticket sizes.

POET's Optical Interposer is a passive Si waveguide platform with hybrid-bonded III-V actives (lasers, modulators, detectors). Architecture-level cost advantage: lower NRE for transceiver-module makers vs. captive monolithic SiPh fabs (Intel, GlobalFoundries Fotonix, Tower SiPho, TSMC SiPh). Validated by 2025 ECOC 'Most Innovative' award (200G/lane Tx/Rx engines, Sept 2025) + 1.6T optical receiver developed with Semtech disclosed at CIOE 2025. Foundry partner: SilTerra Malaysia (April 2018 master collaboration agreement). The architectural pillar of the bull thesis.

Q4 2025 release (Mar 31 2026): management committed to '30,000+ optical engines shipped in 2026' (Venkatesan). At an estimated $300-500 ASP per 800G-class engine, the commitment maps to $9M-$15M of 2026 product revenue — a 9-15× year-over-year increase from the $1.07M FY 2025 base. Anchored by the disclosed $5M production order from 'a leading systems integrator' (Oct 2025) for 2xFR4 + 2xDR4 400G transmit and 800G 2xFR4 + DR8 receive engines, scheduled to ship 2H 2026. This is the first time in POET's history the company has put a unit-volume commitment with a specific year on a public release.

Common bull-thesis framings still describe DenseLight as a current POET subsidiary; primary-source verification corrects this. POET acquired DenseLight (Singapore) in May 2016 and sold it in November 2019 to a Chinese consortium for $26M in cash, retaining only 'preferred supply and strategic cooperation' rights. The post-2019 III-V active stack is rebuilt through partnerships: Sivers Semiconductors (DFB lasers for ELS / CPO; prototypes 1H 2026, production end-2026), NTT Innovative Devices (mobile AI networking; prototypes 2026, high-volume 2027), Semtech (1.6T receiver), Quantum Computing Inc. (3.2T TFLN engines). This is a structurally different model than vertical integration — and is a primary-source ground-truth correction.

December 31 2024: POET completed acquisition of Sanan IC's 24.8% minority equity in Super Photonics Xiamen for total consideration of US$6.5M payable over 5 years (interest free; first payment Oct 31 2025), plus equipment purchase of US$3.8M in four equal installments. The 2020 Sanan IC JV had been framed as POET's production scale-up vehicle for 100/200/400G optical engines — i.e., the strategic transformation play 2020-2024. The $6.5M minority-equity exit price implies Sanan's accounting valued SPX at ~$26M total enterprise value at exit — a fraction of the cumulative capital + management attention deployed. Pattern this establishes: POET enters strategic commitment, commitment underperforms, POET unwinds, new commitment is announced. The current Sivers/NTT/Semtech/QCi partnership stack is the next iteration.

POET discloses partner names (Sivers, NTT, Semtech, QCi) but not module-OEM customer names beyond the generic 'leading systems integrator' framing of the October 2025 $5M production order. Industry-standard NDA practice — same disclosure pattern as Marvell, Coherent, etc. But POET is an order of magnitude smaller, and the implied customer count is much lower. The market-cap-to-EV gap (~$542M cap, ~$110M EV) priced into the equity reflects this opacity: the market is honestly admitting it does not know what the customer-pipeline distribution looks like. Any bull thesis that requires 'trust the pipeline narrative' while the pipeline is unauditable is structurally weaker than a thesis with named customers.

Spot $8.03 close 2026-04-28 × ~67.5M shares O/S = $542M market cap (note: some secondary sources suggest fully-diluted is ~196M shares, requiring FY25 20-F primary-source verification — see open_questions.md Q2). Cash position end-2025 ~$430M + Jan 2026 raise $150M = ~$580M starting-2026 liquidity; less ~$50M payable obligations (SPX + equipment) = ~$430-540M net cash. Implied enterprise value ~$0-110M on a basic-share-count basis, deeply negative on a fully-diluted basis. A simple bull-case math: 30,000 engines × $400 ASP = $12M 2026 revenue → $50-100M 2028 → comparable photonic-component vendors trade at 4-8× EV/Sales for established lines → $250M-$500M incremental EV from product alone. Bull cases compounding several design wins or applying premium AI-photonics multiples (ALAB-style 18-22×) reach $1.5B-$3B EV scenarios.

Co-Packaged Optics is the architectural direction the hyperscaler community is pushing for >1.6T pluggables. Tower Semiconductor's CPO foundry technology announcement (Nov 12 2025) demonstrates merchant-foundry-layer capacity build-out for hyperscaler-direct CPO programs. Marvell's $3.25B Celestial AI acquisition (closed Feb 2 2026) plus the April 22 2026 Polariton POH acquisition creates a scale-up+scale-out optical stack explicitly betting that CPO chiplets become the dominant 2028+ architecture. NVIDIA's announced silicon photonics + CPO program targets vertically-integrated optical interconnect for Blackwell Ultra and Rubin platforms. If hyperscalers move directly to CPO at 1.6T+, the merchant-pluggable-engine TAM compresses materially — the structural threat to POET's bull case Pillar 4 (TAM growth).

POET is currently a Canadian-domiciled foreign private issuer (FPI), filing 20-F annual + 6-K interim reports. Form 4 does NOT apply (Section 16 exempt as FPI). Insider-transaction visibility is materially reduced vs. US-domestic-issuer peers. POET has historically been classified as a Passive Foreign Investment Company (PFIC) for US tax purposes, creating QEF/MTM elections complexity for US holders. POET has signaled US-redomicile evaluation in 6-K disclosures; AGM 2026 (typical June window) is the most likely catalyst window for shareholder vote on redomicile. Completion 2026-2027 would resolve PFIC status and broaden the US institutional investor base — a structural multiple-expansion catalyst.

November 2025: POET announced strategic collaboration with Quantum Computing Inc. (NASDAQ: QUBT) to develop 400G/lane TFLN-modulator-based 3.2 Tbps engines for CPO and AI connectivity. The collaboration extends POET's product roadmap into the 3.2T window where TFLN modulation is increasingly favored by hyperscaler module specifications. Competitive context: TFLN modulators (HyperLight), EO-polymer (Lightwave Logic via the LWLG-Polariton Perkinamine relationship — see LWLG bull case), POH (Polariton/Marvell, acquired April 22 2026) — all compete for the modulator-layer at >1.6T. POET's positioning is at the integration platform layer; the QCi collaboration distributes TFLN IP burden to QCi while POET retains the integration-platform IP.

Four discrete cycles where management commitment to revenue inflection on a 2-year horizon required a 4-6 year horizon (or never materialized at original scale): (1) 2018 SilTerra MCA with 2019-2020 revenue ramp — actual CY2020 revenue ~zero; (2) 2020-2021 Sanan SPX JV with 2022-2023 ramp — JV ended in $6.5M-over-5-years buyout 2024; (3) 2022-2023 100G/400G product cycle with 2023-2024 ramp — actual 2024 revenue $41,427; (4) 2025-2026 800G/1.6T cycle currently in progress with 30,000+ engine commitment for 2026. A reasonable Bayesian prior on the current cycle, conditioned on the prior four, is partial delivery with volume + ASP shading downward, and the next 'decisive transition' cycle requiring a 2027-2028 capital raise.

POET's FY25 annual report on Form 20-F was filed 2026-03-31 (accession 0001493152-26-014253). Headlines: ~$63M FY 2025 net loss; ~$430M cash end-2025; ~$297M accumulated deficit; ~$1.07M FY 2025 NRE + product revenue (vs $41,427 in 2024 — 2,494% YoY growth from a microscopic base). The FY26 20-F is targeted for March 2027 filing window: critical disclosure points will be (a) any material going-concern flag in the auditor opinion, (b) customer-concentration footnote (any > 10% single customer must be disclosed), (c) PFIC determination + US-redomicile track update, (d) risk-factor diff vs. FY25 (added/dropped risks signal management's evolving view of structural threats).