Serenity, the small-cap stock-picker tracked here for his bets on the obscure upstream suppliers feeding the AI buildout, had plenty to react to on Monday — a commercial milestone at his top conviction name and what he read as a widening gap between demand and share prices.

Sivers converts a design-win. The day's headline for him was news that Sivers (SIVE), the tiny Swedish maker of indium-phosphide (InP) CW DFB lasers that is his single largest position, landed its first production orders for continuous-wave lasers used in LiDAR. CW lasers are the light source for silicon photonics; Serenity's whole thesis is that Sivers is the overlooked upstream chokepoint — a critical supplier with few substitutes — for the coming co-packaged-optics (CPO) buildout. He estimated the win could add roughly $20M of blended annual revenue over the next few years at the company's 50-60%+ margins, and stressed the bigger point: it is proof the technology and execution are translating from design-wins into real orders. He flagged the likely customer as Aeva (AEVA), the 4D FMCW LiDAR firm he also owns as a speculative "physical-AI" bet, noting Aeva is reportedly moving into optical sources for hyperscaler CPO/NPO — which would expand Sivers' opportunity. He framed this as one qualification of many still to come, tying it back to AI-datacenter foundry and manufacturing names like GlobalFoundries (GFS), a US silicon-photonics foundry he uses as ecosystem validation, and Jabil (JBL), the contract manufacturer that selected Sivers lasers for its 1.6T pluggable transceivers.

SpaceX earnings as a demand signal. Serenity doesn't own NVIDIA (NVDA) — he treats its roadmap as a telegraph for where upstream demand lands — but he used figures from SpaceX's (SPCX) results to argue the AI-semi cycle is far from over. He pointed to a roughly doubled Q4 capex outlook (into the $35-40B range from a prior ~$18B pace), commentary about deploying many additional gigawatts of power, and the sheer procurement required to stand up hundreds of thousands of GPUs a month. His takeaway: all of that requires enormous volumes of high-bandwidth memory (HBM), power equipment and optical/networking components — the bottlenecks he hunts. SpaceX's blockbuster debut, which he has flagged as a defining mega-IPO, adds to the risk-on backdrop.

Divergence as opportunity. Tying it together, Serenity argued the gap between accelerating fundamental demand and falling equity prices is now at extremes — citing aggressive procurement from Super Micro (SMCI), the AI-server builder he views as cheap after its selloff, and SpaceX, set against collapsing prices in names like Applied Optoelectronics (AAOI), his core Made-in-America optical transceiver long. He suggested October 2026 may later look like the DeepSeek scare or last April's tariff selloff — in other words, a dislocation to buy into rather than flee.

*This is derived commentary for information only, not investment advice.*