Paradis Labs Precise Market Analysis

New Long Idea: OE Solutions (138080.KQ) > OE is one of 4 companies that can make 1.6T laser sources…at just $370M MC??? > And I've not seen anyone mention their name on @X...? > When they're in the same layer on the supply chain as $LITE and $COHR ... but at a fraction of their prices. The core bottleneck is shifting from raw compute cycles -> energy + bandwidth efficiency of the interconnect fabric. OE Solutions is a critical player in the transition from traditional pluggable optics -> CPO. Their CMO said recently: “By providing a high-power, thermally stable light source, we are removing the primary roadblocks to next-generation AI networking.” “We look forward to shipping samples in Q3 2026 to accelerate the global transition to 1.6T infrastructure.” OE Solutions has solved the thermal wall bottleneck that was previously preventing $NVDA & other hyperscalers from moving beyond 800G. The bottleneck has moved to the physical layer of the network. $AVGO etc. have warned that supply chain constraints have migrated to high-power laser components. The move to 1.6T networking requires signaling at 200G per lane. At these speeds, traditional copper fails, forcing a shift to optical fiber for even the shortest distances within the data center rack. The solution is CPO: Which integrates the optical engine directly onto the switch ASIC package. However, CPO introduces a new dependency; the need for an External Laser Source. There are only a couple companies globally capable of manufacturing InP lasers with the necessary power (23dBm) and thermal stability (internal TEC cooling) to drive 1.6T CPO systems. OE Solutions is one of these few. OE Solutions are a Tier-1 supplier to Samsung Electronics and $NOK. They incorporate OE Solutions' transceivers into the networking hardware they sell to hyperscalers like $AMZN, $GOOGL, and $META. The supply chain is something like this: 1 > InP substrates from companies like $AXTI and Sumimoto 2 > InP substrates fabricated into lasers by $LITE, $COHR, OE Solutions 3 > Packaging & photonics by companies like $TSM, $FN, Foxconn/TFC 4 > $NVDA, $AVGO, Samsung, $NOK for networking and system integration 5 > Hyperscalers like $META, $MSFT, $GOOGL $NVDA has direct investments in $COHR + $LITE: But OE Solutions acts as the strategic non-China third source for high-power laser engines needed for the Spectrum-X and Quantum-X photonics platforms. InP laser market: Dominated by three players - $LITE, $COHR, OE Solutions. But OE Solutions has a unique Moat that makes it a superior investment play for the 1.6T cycle: As Western hyperscalers de-risk from Chinese vendors (which are 7/10 globally for modules), OE Solutions is the only vertically integrated Asian player outside of the China-US trade friction. OE Solutions has successfully brought production in-house - while competitors like $COHR are still buying laser chips from $LITE due to yield issues. Internalizing Laser Diode chips allows OE to capture the 50%+ margins that fabless competitors must pay away. They also have a significant software moat: OE's partnership w/ AimValley lets data center operators monitor link health in real-time. A feature $COHR and $LITE typically lack at this price point. In terms of valuation: Historically, OE purchased LD chips externally, keeping gross margins near 10%. W/ internalization, margins are set to expand toward 34% - 42% by late 2026. OE historically traded at a 1.0x - 1.5x P/S as a telecom company. But peer companies like $LITE and $COHR trade at 5x - 8x. Applying a 6.0x P/S multiple to 2027 forecasted revenue of ~$135M: Implies a MC of $804 million USD - More than 2x current. So, undervalued given fwd earnings growth.