Compare moats
Up to three covered companies, band by band. Every call is a curated editorial judgment, never a disclosed figure — and every band carries its cited basis.
| Montage Technology | Nebius Group | GitLab | |
|---|---|---|---|
| Moat rating | narrow The 2026 interim results announcement (published 2026-08-28) cites Frost & Sullivan that in 2024 'the global memory interconnect chip market was highly concentrated, with the top three players accounting for 93.4% of total revenue; we ranked first globally with a market share of approximately 36.8%', and reports H1 2026 gross margin of 65.3%, up 4.87 percentage points. That is a real and profitable position, but its durability is not established: the 2025 annual report (https://www.hkexnews.hk/listedco/listconews/sehk/2026/0424/2026042404088.pdf) warns that 'Early-mover advantages are critical to our ability to capture market share. If competitors launch next-generation products ahead of us during product cycles, we may lose substantial market share', says the DDR5 iteration cycle 'has shortened to 12 to 18 months', and says Samsung Electronics, SK Hynix and Micron Technology 'collectively account for more than 90% of the global server DRAM market'. The lead has to be re-won every sub-generation, in a market whose top three suppliers held 93.4% of 2024 revenue and whose server-DRAM end market three companies hold more than 90% of. And on July 15, 2026 Seoul prosecutors searched the Group's South Korean office in an investigation of a potential competition-law violation (no charges as of the announcement). Narrow, not wide. | narrow The FY2025 20-F positions Nebius as 'one of the few global, at scale, multi-tenant clouds purpose built for AI,' but the same filing names its central dependency plainly: 'We currently rely on Nvidia for the GPU chips we use' — a purpose-built neocloud renting a supply the hyperscalers it competes with also control, which is a real but narrow position. | narrow The FY2026 10-K (year ended 2026-01-31) shows real expansion inside the installed base: 118% Dollar-Based Net Retention, Base Customers up to 10,682 from 9,893, customers above $100,000 of ARR up to 1,456 from 1,229, and $1.0 million ARR customers "to 155 from 123, an increase of 26%". But the filing concedes "limited barriers to entry" and names "Microsoft Corporation, which owns GitHub" as principal competitor. Retention decelerates (130% FY2024, 123% FY2025, 118% FY2026), and since DBNR is reported "on a threshold basis of 130%," the FY2024 figure is a cap and the slide is steeper than it looks. Narrow, not wide: expansion must be re-won against a bundled hyperscaler. |
| Moat type | efficient scale The filing locates the protection in the cost and time of entry, not in exclusive IP: 'Memory interconnect chips feature high entry barriers in terms of R&D capabilities and technical expertise. New entrants require years of intensive R&D to keep pace with the latest industry standards and product iteration cycles, while also completing stringent product validation and qualification processes', and the chips must 'undergo rigorous qualification processes from server CPU vendors and memory modules manufacturers before they can be deployed at scale in commercial applications'. The products are built to open standards ('Leveraging open industry standards, we have established ourselves as a market leader in memory interconnect'), so rivals can and do build compliant parts; what keeps the field small is a market the report's Frost & Sullivan data estimate at USD1,579 million for 2025, in which the top three held 93.4% of 2024 revenue. Montage's role as JEDEC spec owner of DDR5 RCD, MDB and CKD chips is its head start inside that oligopoly, not a lock that excludes the other two. | cost scale The moat the filing describes is engineering: designs that 'optimize power and cooling efficiency, lower latency' plus 'a consistent track record of being one of the first-to-deploy the latest generation of NVIDIA GPU chips' — density and time-to-deploy at scale, not a switching lock. | switching costs Stickiness comes from platform embedding, not IP. The 10-K rests differentiation on "our single platform with a unified data model," on "Consolidation of multiple tools into a single platform," and on deployment inside the customer's own perimeter for regulated buyers. Replacement means re-integrating a toolchain and re-certifying compliance. IP is the wrong label: only "16 issued patents and 25 pending patent applications," plus open-source licenses granting "broad permissions to use, copy, modify, and redistribute." The no-lock-in claim is the filing's own characterisation, footnoted to a Forrester study "commissioned by GitLab." |
| Leadership | co leader Frost & Sullivan, as cited in the interim announcement, ranks Montage first in 2024 memory interconnect chips at approximately 36.8%, with the top three at 93.4%. The two unnamed rivals therefore out-earn Montage combined, and its margin over second place is not disclosed. The company says it 'maintained our global leadership in the DDR5 generation' and calls itself 'one of the only two suppliers of DDR5 Gen 1 MRCD/MDB chips in the world'. In 2024 PCIe Retimers it is second at around 10.9%, in a market where 'the top two companies captured 96.9%'. It is first by revenue in its core market, but inside a three-way oligopoly rather than clear of it. | fast follower The 20-F's own claim is 'one of the FEW global, at scale' AI clouds — a differentiated challenger to the hyperscalers it names as the competitive field, not a claimed leader of it; the first-to-deploy record is a follower's speed advantage, not category leadership. | fast follower The 10-K positions GitLab against a leader rather than as one: "Our principal competitor is Microsoft Corporation, which owns GitHub." Competition reads from the challenger side, differentiating "from GitHub through flexible deployment options that work within enterprise security and compliance requirements, LLM neutrality with self-hosted gateway support" and an open core model. On AI it follows: Duo Agent Platform went "Generally available in January 2026," the last month of the fiscal year. What keeps it a follower rather than behind is cadence, "a new version of our software every month for 172 months in a row," plus a regulated-deployment franchise a SaaS-first rival does not cover. |
| Pricing power | moderate Margins are high and rising. The 2025 annual report (https://www.hkexnews.hk/listedco/listconews/sehk/2026/0424/2026042404088.pdf) gives gross margin of 62.2% in 2025 against 58.1% in 2024, and the interim gives 65.3% for H1 2026, with interconnect chips at 69.3%. But the interim attributes the increase 'mainly' to 'the increase in the proportion of sales revenue of products with higher gross profit margin', which is mix, not price. The buyer side is concentrated. The annual report says the five largest customers took 77.2% of revenue and the largest 29.7%, and it warns that 'significant customer pricing demands or intensified competition, may lead to volatility in our market share and a decline in revenue'. Separately, on July 15, 2026 the Fair Trade Investigation Division of the Seoul Central District Prosecutors' Office began a search and seizure at the Group's South Korean office over 'a potential competition law violation'. Per the Group's July 16, 2026 announcement (https://www1.hkexnews.hk/listedco/listconews/sehk/2026/0717/2026071700057.pdf), none of the Group, its directors or employees had been charged. The Standard (https://www.thestandard.com.hk/finance/article/337444/Montage-Technology-shares-drop-229pc-amid-fair-trade-violation-suspicions, 2026-07-16) reports that prosecutors raided Montage, Renesas Electronics and Rambus 'on suspicion of violating fair trade acts and manipulating prices'. | weak The filing frames a 'highly competitive' market with 'frequent introduction of new or improved solutions' and a GPU cost base set by a single supplier (NVIDIA) — a renter of compute competing on efficiency has little list-price control. | moderate FY2026 gross margin was 87% against 89%, with cost of revenue up $35.6 million "primarily due to an increase of $18.4 million in third party hosting costs for SaaS and cloud usage," and management expects SaaS and Duo mix to bring costs that "may adversely impact our gross margins." Strain is explicit: GitLab "may be required to reduce our prices," competitors "may offer their products and services at a lower price or for free," and it "implemented user limits on our free SaaS product." Power retained: Premium lists at $29 per user/month billed annually, and seats bundle $12 and $24 of GitLab Credits. But "A majority of our subscriptions are on a one-year period." |
| Summary | Montage is a fabless designer of interconnect chips for servers and PCs: DDR5 memory interface chips (RCD/DB), MRCD/MDB for high-bandwidth MRDIMM modules, CKD clock drivers for PC modules, module supporting chips (SPD, PMIC and TS), PCIe Retimers and CXL MXC memory-expander controllers, plus a smaller Jintide x86 server-CPU line built 'With the x86 cores that we procured from Intel'. Its edge is standards leadership in a concentrated niche. It invented the DDR4 '1+9' buffered architecture that was 'ultimately adopted as a JEDEC international standard', it is JEDEC spec owner of DDR5 RCD, MDB and CKD chips, and Frost & Sullivan, as cited in the interim, ranks it first in 2024 memory interconnect chips at approximately 36.8%, with the top three at 93.4%. It extended into PCIe Retimers on in-house SerDes IP and, on the same Frost & Sullivan data, ranked second in 2024 at around 10.9%, as a self-described new entrant. H1 2026 revenue was RMB3,335 million, up 26.7%, with interconnect-chip gross margin of 69.3%. The weaknesses are structural. Per the 2025 annual report, its direct customers for memory chips are module makers in a server DRAM market that three companies hold more than 90% of, and the five largest customers took 77.2% of 2025 revenue. South Korea accounted for RMB2,925 million of RMB5,456 million in 2025 revenue. And on July 15, 2026 the Seoul Central District Prosecutors' Office searched the Group's Korean office in an investigation of a potential competition-law violation. | Nebius sells a 'unified full-stack AI cloud that spans the complete AI journey – from compute capacity to software and services,' with hardware and software built in-house — the neocloud pitch of hyperscaler reliability at purpose-built efficiency, and a first-to-deploy record on new NVIDIA silicon. Two things bound the moat, both from the filing: it 'currently rel[ies] on Nvidia for the GPU chips,' the same constraint every neocloud shares, and it is a Nasdaq 'Controlled Company' whose founding shareholder holds concentrated voting power. The three non-core segments (Toloka, Avride, TripleTen) are separate businesses, not the cloud moat. | GitLab sells a single-platform DevSecOps toolchain defended by workflow lock-in rather than intellectual property: a unified data model plus deployment flexibility, including single-tenant "Dedicated for Government with FedRAMP compliance," makes it hard to remove where data residency is contractual. The installed base expands: 118% Dollar-Based Net Retention, 10,682 Base Customers, 1,456 above $100,000 of ARR, 155 above $1.0 million (up 26%), over 70% of ARR from public sector and enterprise. The filing supplies the counterweight: "limited barriers to entry," Microsoft/GitHub as principal competitor, 16 issued patents. Retention (130 to 123 to 118, the 130 a cap), gross margin (89% to 87%) and a January 2026 AI launch point to compression at the commodity end. No source measures market share, so every share band is "unknown". |
| Chain position | Upstream fabless chip supplier. Per the 2025 annual report, memory interface and supporting chips go to memory module manufacturers ('our direct customers being memory module manufacturers'), and PCIe Retimers go to server OEMs/ODMs as direct customers, with cloud providers as end users. Wafer fabrication, packaging and testing are outsourced, and the five largest suppliers were 79.4% of 2025 purchases. | Layer-8 neocloud — a purpose-built AI compute provider reselling NVIDIA silicon at scale. | GitLab sits at the software-tooling layer, not compute or models. It consumes third-party LLM capacity — marketing "LLM neutrality and support for self-hosted AI gateways, including air-gapped environments" — so it captures no model-layer economics and bears inference as COGS, visible in the $18.4 million rise in hosting costs. Its distinctive position is the regulated perimeter. |
| Products (share / barrier) |
|
|
|
| Long-horizon vote | +0.13 at weight 0.20 · swarm neutral Editorial prior, not backtested. | -0.01 at weight 0.20 · swarm neutral Editorial prior, not backtested. | +0.06 at weight 0.20 · swarm neutral Editorial prior, not backtested. |