护城河对比
最多三家已覆盖的公司，逐项对照。每一项判断都是人工编辑的判断，而非公司披露的数字——并且每一项都附有其引用依据。
| Marvell Technology | Cerebras Systems | Wiwynn | |
|---|---|---|---|
| 护城河评级 | 窄 FY2026 10-K (filed 2026-03-11): differentiated platform IP — over 10,000 issued patents and pending applications as of 2026-01-31, plus a proven custom ASIC platform leveraging ultra-high-speed SerDes, silicon photonics, co-packaged optics and custom HBM — but Marvell itself calls its markets 'intensely competitive' with 'pricing pressures', notes customers 'have chosen to develop certain semiconductor products internally', and discloses two >=10% customers with the ten largest at 82% of FY2026 net revenue. Real, defensible IP in a concentrated, contestable customer base = narrow, not wide. https://www.sec.gov/Archives/edgar/data/1835632/000183563226000011/mrvl-20260131.htm | 窄 One-of-a-kind silicon sold to very few buyers: the IPO prospectus discloses G42 at 24.0% of 2025 revenue (85.0% in 2024), MBZUAI at 62.0% of 2025 revenue, and a December 2025 MRA with OpenAI representing 'a substantial portion of our projected revenues over the next several years' — against a competitive field the same document names as NVIDIA, AMD, Intel, hyperscaler accelerators and the major clouds. | 窄 The 2025 annual report (published 2026-04-30) shows a real but shallow position: products deployed in "more than 750 hyperscale data centers" and, on the Company's own internally assessed shipment volume, "more than 10% of global server shipments and nearly 30% of global ODM-Direct server shipments" in 2025 — while the same report's own list of negative factors names "Excessive concentration of customers" and a market "getting more and more competitive", discloses three customers at 41.48%, 26.03% and 10.43% of 2025 net sales, and reports 2025 gross profit up 109.77% against revenue up 163.68% for an 8.3% gross margin. Defensible incumbency, not a durable one. |
| 护城河类型 | 无形资产与知识产权 The moat rests on hard-to-replicate mixed-signal IP: the 10-K describes the custom ASIC platform built on ultra-high-speed SerDes, ARM compute, security, storage, silicon photonics and advanced packaging (die-to-die interconnects, chiplets, CPO, custom HBM), with multiple 5nm designs executed, 3nm in progress and a 2nm platform in development; a secondary switching-cost element comes from multi-year custom design wins co-developed to individual customer specifications. https://www.sec.gov/Archives/edgar/data/1835632/000183563226000011/mrvl-20260131.htm | 无形资产与知识产权 Wafer-scale integration is the moat's substance: the prospectus contrasts the WSE-3's 4 trillion transistors on 46,225 mm² of silicon with NVIDIA's B200 at 208 billion on roughly 1,600 mm² — a commercial chip category with one occupant. | 转换成本 The report's "Competitive niche" section locates the advantage in the customer program rather than in IP or unit cost: "highly customized solutions for hyperscale data centers", "end-to-end technical integration capabilities ranging from server design and system integration to large-scale L10/L11 rack-level delivery", and a one-stop service "covering product design, integration, optimization, deployment, and after-sales services" across a base of more than 750 hyperscale data centers. Every advantage the section claims is about integration work and customer proximity, none about patents or proprietary silicon, and an 8.3% gross margin is not the profile of a cost advantage; what is left is the friction in an installed, co-designed rack program. |
| 领先地位 | 并列领先者 Leader in its optics niche, #2 in custom silicon: ~60% of high-end PAM4 DSP share (36kr, 2026-06-27, https://eu.36kr.com/en/p/3870758441178373) but an estimated 20-25% of custom AI ASIC design services versus Broadcom's ~70% (hashrateindex, 2026-05-13, https://hashrateindex.com/blog/design-partners-ai-asic-market-part-2/) — net, a co-leader in AI data-center connectivity/custom silicon behind Broadcom overall. | 明确领先者 Scoped to the category the record supports: wafer-scale integration has no other commercial occupant per the prospectus's own comparison, and the Q2 release claims — attributed, via AMD and AWS partnerships — leadership in disaggregated inference, a technique the prospectus defines. | 并列领先者 The report claims "nearly 30% of global ODM-Direct server shipments" for 2025 on internally assessed volume, against an IDC Worldwide Quarterly Server Tracker base of approximately 7.7 million ODM-Direct units within roughly 16.78 million total 2025 server shipments — yet describes the Company only as "one of the major suppliers to international hyperscale data centers". Among the largest in its segment; the report makes no claim to be first. |
| 定价权 | 中等 The 10-K characterizes Marvell's markets as having 'pricing pressures' and intensifying competition (https://www.sec.gov/Archives/edgar/data/1835632/000183563226000011/mrvl-20260131.htm), yet the Q1 FY2027 release reports 52.1% GAAP / 58.9% non-GAAP gross margin on record revenue (https://www.sec.gov/Archives/edgar/data/1835632/000183563226000014/q127_8kx522026ex-991.htm) — differentiated-IP margins, tempered by hyperscaler buyer power. | 中等 The CFO states Q2 'significantly improved core gross and operating margins compared to a year ago' — improving, from a base where three named customers hold most of the revenue and the leverage that comes with it. | 弱 2025 gross margin was 8.3% and operating margin 6.7%; gross profit rose 109.77% while revenue rose 163.68%, so the margin narrowed as volume grew. The report discloses three customers at 41.48%, 26.03% and 10.43% of 2025 net sales, lists "Excessive concentration of customers" among its negative factors, and explains its main-supplier table with "The suppliers are designated by customers." |
| 综述 | Marvell is a fabless data-infrastructure silicon supplier whose center of gravity has shifted decisively to the AI data center: the data center end market was $6,100.3M, 74% of FY2026 revenue, up from 40% two fiscal years earlier (FY2026 10-K). Its strongest position is electro-optics — in high-end PAM4 optical DSPs for 400G+ transceivers it holds roughly 60% share on Inphi-inherited SerDes/FEC IP, with Broadcom above 30%, the two together over 90% (36kr, 2026-06-27). In custom AI silicon it is the structural #2 design partner at an estimated 20-25% of the custom AI ASIC design-services market versus Broadcom's ~70%, anchored by AWS Trainium and Microsoft Maia wins (hashrateindex, 2026-05-13). The Q1 FY2027 release (2026-05-27) shows the flywheel turning — record $2.418B revenue (+28% YoY), Q2 guided to $2.7B mid-point (+35% YoY), management citing 'exceptional AI-related bookings' across 800G/1.6T optics, 51.2T switches, CPO/NPO and custom XPU — and the Celestial AI (Photonic Fabric) and XConn (PCIe/CXL switching) acquisitions closed in February 2026 extend the interconnect moat toward scale-up fabrics. The offsets that keep the moat narrow are in Marvell's own filing: intense competition (AMD, Alchip, Astera, Ayar, Broadcom, Credo, GUC, Lightmatter and others), hyperscaler in-housing risk, and heavy customer concentration. | Cerebras bet that AI compute should be one enormous chip rather than many small ones — 'quantities of compute and memory never before assembled on a single commercial chip,' avoiding 'the latency and the power-draw induced by the traditional approach,' in the prospectus's words — and is converting that bet into an inference cloud: Q2 2026 cloud revenue grew 281% with 600 MW of data center capacity under contract and $25.4 billion of remaining performance obligations. The concentration that funded the bet is the risk that remains: two related-party Gulf customers were 86% of 2025 revenue between them, with the OpenAI MRA now layered on top. | Wiwynn sells ODM-Direct — it designs and builds cloud and AI server systems to hyperscale operators' specifications rather than shipping a branded product — and the 2025 annual report grounds its edge in that relationship: "highly customized solutions", integration "from server design and system integration to large-scale L10/L11 rack-level delivery", Open Compute Project participation described as "a core competitive advantage in marketing to global hyperscale data centers", and products deployed in more than 750 hyperscale data centers as of the end of 2025. The report also says that high integration and testing abilities "increases the threshold of vendors in the assembly industry and creates entry barriers". Against that, the same report lists excessive customer concentration and an increasingly competitive market as its negative factors, discloses customers at 41.48%, 26.03% and 10.43% of 2025 net sales, explains its main-supplier table with "The suppliers are designated by customers", and reports an 8.3% gross margin on 2025 revenue of NT$950.66bn. |
| 产业链位置 | Fabless supplier spanning 'data center core to network edge': it sits between hyperscaler AI compute (custom XPU/XPU-attach ASICs) and the optical layer (PAM4/coherent DSPs, CPO/LPO, DCI, AEC, PCIe retimers), outsourcing fabrication to independent CMOS foundries; the Feb-2026 Celestial AI and XConn acquisitions push it further into scale-up photonic fabric and PCIe/CXL/UALink switching (FY2026 10-K, https://www.sec.gov/Archives/edgar/data/1835632/000183563226000011/mrvl-20260131.htm). | Layer-4 wafer-scale accelerators plus an inference cloud — hardware vendor and specialized cloud in one node. | The report's cloud/AI-server value-chain diagram places system assembling in the midstream and "Direct sales from original manufacturers" downstream selling to cloud service providers; Wiwynn describes itself as providing "IT equipment and rack system integration service to hyperscale data centers". 82.12% of 2025 revenue came from America. |
| 产品（份额 / 壁垒） |
|
|
|
| 长周期投票 | +0.13 权重 0.20 · 群体判断 bullish 编辑先验，未经回测。 | +0.20 权重 0.20 · 群体判断 bearish 编辑先验，未经回测。 | +0.05 权重 0.20 · 群体判断 bearish 编辑先验，未经回测。 |