Deep Research Alpha Scan — 2026-03-15
Scan mode: abbreviated frontier tech and pre-breakout scan. Full analysis deferred due to context budget.
Methodology
Scoring: /10 scale across technology readiness, commercial traction, market size, competitive moat, and timing. Only opportunities scoring 6+ with "Now" or "Watch" timing highlighted.
Frontier Tech Opportunities
| # | Topic | Company/Ticker | Score | Timing | Upside |
|---|---|---|---|---|---|
| 1 | Silicon anode batteries for defense drones | Amprius (AMPX) | 8/10 | Now | 3-5x over 24 months |
| 2 | AI cloud infrastructure (NVIDIA ecosystem) | Nebius (NBIS) | 7/10 | Now | 2-3x over 18 months |
| 3 | Advanced fission microreactors | Oklo (OKLO) | 7/10 | Watch | 5-10x over 5 years |
| 4 | Silicon photonics / co-packaged optics | Multiple (see below) | 7/10 | Watch | Sector-level opportunity |
| 5 | Telehealth + branded GLP-1 distribution | Hims & Hers (HIMS) | 7/10 | Now | 50-100% over 12 months |
Detailed Assessments
1. AMPX — Silicon Anode Batteries (8/10, Now)
Technology readiness: Commercially shipping. 520 Wh/kg cells recognized at CES 2026. First positive EBITDA quarter achieved.
Commercial traction: Nokia drone-in-a-box contract. Defense Replicator Program creating guaranteed demand for high-endurance batteries. Nanotech Energy manufacturing partnership secures domestic production.
Why now: Revenue grew 202% YoY to $73M. 2026 guidance exceeds $125M. The defense drone market is accelerating toward $40B+. Pentagon's Drone Dominance program placing first orders. AMPX is one of very few companies with flight-qualified silicon anode cells.
Risk: Pre-profitability on GAAP basis. Manufacturing scale-up execution risk. Competition from CATL, Samsung SDI on energy density.
2. NBIS — AI Cloud Infrastructure (7/10, Now)
Technology readiness: Operational GPU clusters. Missouri 1.2 GW campus approved.
Commercial traction: NVIDIA $2B investment with co-development. Early access to Rubin platform, Vera CPUs, BlueField storage. 5 GW capacity target by 2030.
Why now: The NVIDIA deal transforms NBIS from "another GPU cloud" to a co-development partner with early access to next-gen hardware. This is the kind of strategic relationship that created the hyperscaler moats.
Risk: Execution on multi-GW buildout. Capital intensity. Competition from established clouds (AWS, Azure, GCP).
3. OKLO — Advanced Fission Microreactors (7/10, Watch)
Technology readiness: Pre-construction. NRC environment increasingly favorable (TerraPower got permit in 18 months, 9 months ahead of schedule).
Commercial traction: Meta prepayment deal for Ohio plant. KHNP (Korea Hydro & Nuclear Power) MOU for international SMR development. Business update March 17.
Why now: The Iran war has made energy security a national priority. Trump administration wants to quadruple US nuclear capacity from 100 GW to 400 GW by 2050. NRC is approving faster than ever. But OKLO is still pre-revenue — this is a years-long thesis.
Risk: No revenue. NRC timeline uncertainty. Fuel supply (HALEU availability).
4. Silicon Photonics (7/10, Watch)
Key developments: TSMC's SiPh manufacturing maturing for co-packaged optics. Optical transceiver market projected to exceed $12B by 2026. Samba Photonics achieved 95% coupling efficiency. MIT developed efficient free-space photonic beam device.
Public exposure: Broadcom (AVGO), Marvell (MRVL), Lumentum (LITE), Coherent (COHR). Early-stage: Lightmatter (private).
Why now: AI data center bandwidth demands are hitting copper wall. CPO is the answer and 2026-2027 is the commercialization inflection.
5. HIMS — Telehealth GLP-1 Distribution (7/10, Now)
Commercial traction: Novo Nordisk partnership announced March 9. Patent lawsuit dismissed. FDA Commissioner endorsed deal model.
Why now: This is a thesis-changing event. HIMS went from potential patent infringement defendant to authorized Novo Nordisk distributor overnight. The GLP-1 market is projected to exceed $100B annually. HIMS now has a legitimate, legal, FDA-endorsed pathway to participate.
Risk: Deal terms may compress margins vs. compounding. Novo can refile lawsuit. Competitive telehealth landscape.
Below Threshold
| Topic | Score | Why Not Now |
|---|---|---|
| Graphene commercialization | 5/10 | Still 3-4 years from scale. Samsung display plans 2026-2027 but no pure-play public stock. |
| Quantum computing | 5/10 | IBM blueprint encouraging but commercial applications still limited. INFQ too early. |
| Solid-state batteries | 5/10 | Toyota/CATL targeting 2027 prototypes. GAC pilot line active. No public pure-play at right stage. |
| Humanoid robotics | 5/10 | Tesla Optimus Gen 3 in production but external sales not until late 2026. "Digital Optimus" with xAI is vaporware until proven otherwise. |
Generated: 2026-03-15 | Scan mode — abbreviated scoring based on available data