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.
| Kioxia Holdings | Nebius Group | NVIDIA | |
|---|---|---|---|
| Moat rating | narrow Kioxia's FY2025 Annual Securities Report (dated 2026-06-24) describes a real but bounded position. It records 'the limited number of major competitors in the flash memory business', and says its three manufacturing joint ventures with Sandisk let it 'make investments on a larger scale than if it were to invest alone and enjoy economies of scale in terms of capital expenditures and production efficiency'. Against that, the same Risk Factors section describes 'heavy global competition as advanced technologies are necessary for business execution', says 'in the flash memory market the Group's ability to decide prices is limited', that 'Some competitors possess technologies that the Group does not have, such as DRAM' and 'have greater financial strength than the Group', and records that the Group cut production from October 2022 to March 2024. Scale in a small field, not a barrier that protects Kioxia's returns through the cycle. | 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. | wide Wide: the CUDA software platform and its developer installed base create high switching costs across the AI-training stack, defended by $76.7B cumulative R&D and reinforced by ~71% FY2026 gross margins and Data Center revenue (~90% of total) up 68% YoY — durable, hard-to-replicate advantages per the FY2026 10-K. |
| Moat type | cost scale The advantage the company itself names is scale and capital efficiency. The Annual Securities Report says the Sandisk joint ventures, which procure the production equipment installed at Kioxia's Yokkaichi and Kitakami plants and 'sell 50% of their products to the Group and 50% to the Sandisk Group', let it 'enjoy economies of scale in terms of capital expenditures and production efficiency'. At its 2026 Investor Day (2026-06-02) management said that 'by leveraging our world-leading economies of scale, we have maintained a lower cost per gigabyte than the industry average', on a slide that cites the TechInsights NAND Market Report Q2 2026 as its source. The filing describes its technology work (layer stacking for BiCS FLASH, the move to QLC) and names 'a decline in competitiveness in production efficiency per gigabyte' as a risk if it falls behind. It does not present that technology as protected IP that rivals cannot match. | 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. | intangibles ip The durable edge rests on proprietary IP and software: the full-stack CUDA development platform running on all NVIDIA GPUs plus hundreds of proprietary domain libraries/SDKs/APIs, and $76.7B cumulative R&D yielding "inventions that are essential to modern computing" (NVIDIA invented the GPU in 1999). This is reinforced by developer-ecosystem network effects and CUDA switching costs, per the FY2026 10-K Business section. |
| Leadership | at parity TrendForce's 2Q26 NAND Flash ranking (2026-08-18) puts Kioxia fourth with revenue of about $10.72 billion. Its 13.6% share 'edged down', and it sits behind Samsung (29.3%), SK hynix Group and Micron, which moved up to third. TrendForce's 2Q26 enterprise SSD ranking (2026-09-01) also places it fourth, at $4.64 billion. On technology it is keeping pace: it began sampling the 332-layer 10th-generation BiCS FLASH on 2026-07-03, per its press release. The Annual Securities Report concedes that some competitors 'possess technologies that the Group does not have, such as DRAM' and 'have greater financial strength than the Group'. Kioxia is in the middle of a small group of peers. | 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. | clear leader FY2026 Data Center revenue was $193.7B (~90% of the $215.9B total), up 68% YoY on the Blackwell ramp, and NVIDIA describes itself as "a data center scale AI infrastructure company reshaping all industries." The 10-K frames named rivals as parties who "provide or intend to provide" GPUs/accelerators — incumbent-leader positioning. |
| Pricing power | weak The Annual Securities Report states that 'in the flash memory market the Group's ability to decide prices is limited, and in the medium- to long-term selling prices on a bit basis may decline at a similar pace as in the past'. Current margins follow the market. The first-quarter FY2026 results attribute the revenue jump primarily to 'a significant increase in average selling prices (ASPs) resulting from strong demand from data center customers focusing on generative AI'. Gross profit was ¥1,380,066 million on revenue of ¥1,767,117 million, against ¥71,179 million on ¥342,799 million a year earlier. In the 2026 Investor Day Q&A management said it believes 'market pricing will continue to reflect those supply-demand conditions'. | 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. | strong FY2026 gross margin was 71.1% (75.0% in FY2025); per the 10-K MD&A the ~3.9pt decline reflects the Hopper HGX→Blackwell full-system mix shift and a one-time $4.5B H20 excess-inventory/purchase-obligation charge, not competitive price erosion. A low-70s% hardware gross margin evidences strong pricing power. |
| Summary | Kioxia calls itself 'a specialized flash memory manufacturer'. It makes BiCS FLASH 3D NAND at Yokkaichi and Kitakami in Japan through three manufacturing joint ventures with Sandisk, whose output is sold 50% to Kioxia and 50% to Sandisk. FY2025 (year to March 2026) revenue was ¥2,337.6bn: SSD & Storage ¥1,362.6bn, Smart Devices ¥760.0bn and Other ¥215.0bn, with the Apple group alone at 20.4% of sales. Its moat is scale in a concentrated industry. TrendForce ranks it fourth in 2Q26 NAND revenue with a 13.6% share, and in August 2026 the partners announced anticipated investments in Japan of over $31 billion through 2032, contingent on government support. The same filing sets out the limits: limited ability to set prices; rivals that also make DRAM and have more financial strength; China's government-supported push for domestic semiconductor production; and a joint-venture agreement effective until 2034, with no decision yet on whether it continues after that. First-quarter FY2026 gross profit was ¥1,380,066 million on revenue of ¥1,767,117 million, against ¥71,179 million on ¥342,799 million a year earlier. The company attributes the jump primarily to a significant increase in average selling prices. | 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. | NVIDIA pairs market-leading accelerated-computing hardware (the Blackwell data-center platform) with a proprietary full-stack software moat — CUDA plus hundreds of domain libraries — funded by $76.7B of cumulative R&D, and its "large and growing number of developers and installed base... strengthens our ecosystem and increases the value of our platform for our customers" (FY2026 10-K). Competition is intensifying from AMD, Intel and Huawei, and from hyperscalers (Alphabet, Amazon, Microsoft) designing internal AI silicon, but rivals must overcome NVIDIA's entrenched CUDA software ecosystem to displace it. |
| Chain position | Upstream NAND flash and SSD supplier. The filing says much of its revenue 'depends on a limited number of customers and industries, such as major smartphone manufacturers and large IT companies, including hyperscalers that require SSDs'. The Apple group was 20.4% of FY2025 sales. | Layer-8 neocloud — a purpose-built AI compute provider reselling NVIDIA silicon at scale. | Upstream compute-platform supplier: NVIDIA sells full-stack data-center systems (GPU + Arm CPU + DPU + NVLink/InfiniBand networking + CUDA software) to cloud providers and enterprises, and relies on third-party foundry/assembly-test-packaging partners (e.g., SPIL, Amkor, Wistron, Fabrinet). Per the FY2026 10-K, several of its largest customers (hyperscalers such as Amazon, Alphabet, Microsoft) are simultaneously customers and emerging competitors developing internal accelerated-computing silicon. |
| Products (share / barrier) |
|
|
|
| Long-horizon vote | -0.01 at weight 0.20 · swarm neutral Editorial prior, not backtested. | -0.01 at weight 0.20 · swarm neutral Editorial prior, not backtested. | +0.42 at weight 0.20 · swarm bullish Editorial prior, not backtested. |