Chapter 13 / 16
Core, Watch, and Warning Pools
核心、观察、拥挤赢家、等待回调、低共识瓶颈与叙事回避必须显式分池。
Current Qualitative Consensus Map
As of 2026-06-20, based on recent market and industry signals:
| Theme | Bottleneck | Consensus / priced-in | Discovery gap | Notes |
|---|---|---|---|---|
| Nvidia GPU / full-stack AI compute | 5 | 5 | 0 | Real bottleneck, but fully consensus. Use as demand signal, not hidden opportunity. |
| HBM leaders: SK hynix, Micron, Samsung | 5 | 4-5 | 0-1 | Very real, but now widely recognized. Still cyclical upside possible, less undiscovered. |
| Nearline HDD: Seagate, Western Digital | 4 | 3-4 | 1 | AI storage demand is being discovered; not hidden anymore, but less crowded than HBM. |
| CXL / memory pooling: Astera, Rambus, 澜起科技 | 4 | 2-3 | 1-2 | Better non-consensus layer: key if memory shifts from fixed server DRAM to pooled memory. |
| DDR5 interface / retimer: 澜起科技, Rambus, Astera | 4 | 2-3 | 1-2 | More attractive than generic memory modules because it sits in the control/interconnect layer. |
| Enterprise SSD / NAND controllers: Silicon Motion, SanDisk, 江波龙, 佰维存储 | 3-4 | 3 | 0-1 | Good cycle trade, but more exposed to commodity pricing and inventory swings. |
| AI host CPU attach: AMD, Arm, Nvidia Grace/Vera | 4 | 3-4 | 0-1 | Becoming visible; still worth tracking because GPU rack architectures force CPU attach. |
| Intel CPU / foundry / packaging turnaround | 3-4 | 3 | 0-1 | Potentially large, but execution risk is high; separate turnaround thesis from AI bottleneck thesis. |
| A-share国产CPU: 海光信息, 龙芯中科 | 3-4 | 3-4 | 0-1 | More policy/import-substitution than pure AI bottleneck; valuation often already anticipates this. |
| AI server OEM/ODM: 浪潮信息, 中科曙光, 工业富联 | 3 | 3-4 | -1-0 | Demand is obvious, margins may be competed away; watch order quality and profit capture. |
Automation Label
Each company should receive one of these labels:
crowded_winner: real winner, but already heavily priced.watch_for_pullback: good business, weak margin of safety at current consensus.underdiscovered_bottleneck: best target for deep research.cycle_beta: mostly memory / hardware cycle exposure.execution_option: upside depends on company execution, not only industry bottleneck.avoid_narrative_only: AI story exists, but bottleneck capture is unclear.
Initial labels:
| Company | Market | Label | Rationale |
|---|---|---|---|
| Nvidia | US | crowded_winner | Strongest AI infrastructure position, but fully consensus. |
| Micron | US | watch_for_pullback | HBM/DRAM/NAND bottleneck is real, but memory boom is now widely recognized. |
| Seagate | US | watch_for_pullback | Nearline HDD scarcity is increasingly recognized; still less crowded than HBM. |
| Western Digital | US | watch_for_pullback | Similar HDD/storage thesis; monitor pricing discipline and supply commitments. |
| Astera Labs | US | underdiscovered_bottleneck | CXL/retimer/connectivity layer is strategically important, but valuation may already embed some scarcity premium. |
| Rambus | US | underdiscovered_bottleneck | Memory-interface IP and chips may benefit from DDR5/CXL without taking full commodity memory risk. |
| AMD | US | watch_for_pullback | CPU attach and AI accelerator upside are real, but increasingly consensus. |
| Arm Holdings | US | watch_for_pullback | Server Arm attach is real, but valuation often prices a broad architecture win. |
| Intel | US | execution_option | Foundry/packaging/CPU recovery has upside, but thesis depends on execution. |
| 澜起科技 | A-share | underdiscovered_bottleneck | DDR5 interface, PCIe retimer, CXL MXC; better positioned in memory interconnect than generic storage names. |
| 海光信息 | A-share | watch_for_pullback | Domestic CPU leader; strong policy/国产替代 narrative already partly priced. |
| 中科曙光 | A-share | cycle_beta | AI server/system demand beneficiary; profit capture needs verification. |
| 浪潮信息 | A-share | cycle_beta | High AI server relevance, but more competitive hardware margin profile. |
| 江波龙 | A-share | cycle_beta | Storage cycle and module upside, but more exposed to commodity swings. |
| 佰维存储 | A-share | cycle_beta | Similar storage cycle beta; needs evidence of durable enterprise/AI mix. |
Primary automation sort:
research_priority = monopoly_potential_score + discovery_gap + evidence_velocity - valuation_heat
Automation should surface:
- New high-priority company mentions.
- Companies with rising evidence velocity.
- New bottleneck terms that appear across multiple sources.
- Disagreements between investor narrative and operational evidence.
- Companies where consensus score rises faster than evidence velocity, which may mean the trade is becoming crowded.
Observation Pool, Not Core Pool
- SIVE / Sivers Semiconductors: high-signal photonics bottleneck candidate, supported by Jabil and GF collaborations, but social heat and valuation expectations are high. Need independent proof of share, gross margin, capacity rights, and product revenue ramp.
- AAOI / Applied Optoelectronics: repeatedly appears in Serenity-derived photonics discussions, but needs a narrow monopoly layer beyond "optical networking demand."
- POET / Celestial AI / Ayar Labs ecosystem names: strategically important to CPO / optical compute narratives, but public equity capture and supplier economics differ by company.
- PENG / Penguin Solutions: useful CXL product evidence; core inclusion requires proof that Penguin captures margin versus being a systems integrator.
- MRVL / Marvell: strong AI connectivity and custom silicon narrative; likely more consensus now. Treat as platform winner/watch_for_pullback unless a specific unavoidable layer is isolated.
- NBIS / neocloud and powered data-center names: contracted power is valuable, but equity value depends on active MW, utilization, customer credit, and financing terms.
- 绿的谐波: humanoid/physical AI component thesis is plausible, but this week's move looks social-liquidity driven; separate from AI data-center radar.
Warning Pool
- Any Serenity-mentioned A/H stock that gaps before primary evidence appears. 21jingji's 绿的谐波 report is a case study in social heat causing immediate price action.
- SIVE if the thesis is reduced to "Serenity likes it" or if market cap prices in 20-30x revenue growth before volume revenue is visible.
- Generic storage names after broker target hikes and media stories about sold-out HDDs. Real shortage, but discovery_gap is shrinking.
- AI data-center power names with only headline GW/MW claims and no interconnection, PPA, permitting, or active-load evidence.
- CXL names where the product is technically interesting but no design win, attach-rate, or ASP/margin evidence exists.