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.
| IREN | Tempus AI | AMD | |
|---|---|---|---|
| Moat rating | none The FY2026 10-K (filed 2026-08-27) shows IREN holds an input the industry competes on: it names 'access to secured and energized power' first among what it believes are the principal competitive factors in its industry, and reports 'executed grid connection agreements, letters of agreement or equivalents representing approximately 5GW of total power capacity' in the United States, Canada, Spain and Australia as of June 30, 2026. It does not claim an advantage over rivals in that input. It says 'Certain competitors may have access to more competitively priced power, a greater access to power and a better capacity to timely secure grid connections,' and 'Many of our competitors have greater financial, technical or commercial resources, longer cloud operating histories or larger customer bases, compared to us.' Operating AI Cloud Services capacity was approximately 40MW at June 30, 2026, and the stored XBRL fundamentals from this 10-K show an FY2026 operating loss of about $1,046.7M on revenue of about $707.0M, so no durable advantage is yet demonstrated. | narrow The FY2025 10-K documents a real asset: a network "connected in some way to more than 55% of all oncologists practicing in the United States" and "more than 65% of all academic medical centers," via "more than 700 unique data connections" across "more than 5,000 healthcare institution sites," with "broad data rights, including the rights to longitudinally updated data." Not eroding: revenue $693.4M to $1,271.8M (+83%), net loss narrowed from $705.8M to $245.0M. Not wide: still loss-making against "an accumulated deficit of $2.4 billion," with payment received on only "approximately 55% of our clinical oncology NGS tests and 50% of our hereditary tests." | narrow AMD's FY2025 10-K documents real but bounded advantages: approximately 7,200 U.S. patents and roughly 18,900 patent matters worldwide, an outright claim that 'We are the market share leader in semi-custom game console products,' and gross margin of 50% on revenue up 34% to $34.6 billion. The same filing sets the ceiling: 'Some of our competitors may possess stronger market positions, larger customer bases, more design wins, and greater financial, sales, marketing, and distribution resources than us'; Nvidia is named 'the discrete GPU market share leader'; and AMD relies on TSMC 'for the production of all wafers for microprocessor and GPU products at 7 nanometer (nm) or smaller nodes.' Advantages that are genuine, contested, and dependent on a shared foundry are narrow, not wide. |
| Moat type | none The FY2026 10-K's asserted advantage is control from owning its data centers, 'including the associated land, grid connections and substations', which it believes allows it 'to benefit from more sustainable cash flows and operational flexibility relative to operators that rely upon third-party colocation services or short-term land leases'. Its only scale language is a belief that its 'procurement scale, deployment experience and direct control over the data center layer' lets it bring new compute into service 'rapidly and at scale', a deployment-speed claim, and it 'generally target[s]' regions with 'low-cost and attractive renewable energy sources', a siting target rather than a demonstrated cost position. It does not claim a cost or scale lead: it says 'Many of our competitors are larger, have longer operating histories and significantly greater resources than we do' and that certain competitors 'may have access to more competitively priced power, a greater access to power and a better capacity to timely secure grid connections.' There is no network effect or lock-in either, since existing contracts have 'terms ranging from month to month up to five years'. No enumerated moat source is grounded. | network effects The filing states the mechanism: "Each product line is designed to enable and enhance the other, thereby creating network effects." That is the filing's own characterisation, so the rating rests on corroborating facts: 700+ data connections across 5,000+ sites with longitudinal rights; roughly 801,000 tests in FY2025 (about 340,500 Oncology) feeding the records Insights licenses; Insights working with "19 of the 20 largest public pharmaceutical companies based on 2024 revenue" at "approximately 126%" net revenue retention. Intangibles fit less well: the xT PMA, ADLT status, Paige's clearance and five CAP/CLIA labs are real, but named peers hold comparable approvals. | intangibles ip The FY2025 10-K locates the durable asset in design IP rather than manufacturing: 'We rely on contracts and intellectual property rights to protect our products and technologies from unauthorized third-party copying and use,' with approximately 7,200 U.S. patents and about 18,900 patent matters worldwide spanning x86 EPYC/Ryzen CPUs, CDNA/RDNA graphics and the Versal/Zynq adaptive-SoC families, and it credits customer wins to 'our broad IP portfolio and leadership in design, integration and advanced packaging.' It is explicitly not cost_scale: AMD is fabless and 'utilize[s] Taiwan Semiconductor Manufacturing Company Limited (TSMC) for the production of wafers' — the same leading-edge foundry its merchant competitors use, so no manufacturing-scale advantage accrues to it. |
| Leadership | behind The 10-K lists IREN's competitors as hyperscalers (Amazon Web Services, Google Cloud, Microsoft Azure, Oracle Cloud) and specialized AI Cloud Services providers (CoreWeave, Nebius, Crusoe, Lambda, Nscale, SpaceX and others). It says IREN began providing AI Cloud Services in 2024 and cites its 'shorter operating history in AI Cloud Services relative to some competitors', and states 'Some of our competitors have longer operating histories, larger customer bases, more comprehensive IP portfolios and patent protections, more design wins, and greater financial, sales, marketing and distribution resources than we do.' Operating AI Cloud Services capacity was approximately 40MW at June 30, 2026. The filing describes a smaller, later entrant that trails some competitors on operating history, customer base and resources, with its progress so far resting on two anchor contracts (a ~$9.7B five-year Microsoft agreement and a ~$3.4B five-year NVIDIA contract) rather than any claimed category position. | at parity The filing claims no ranked lead; its positional language is hedged, as in "one of the largest sequencers of cancer patients... in the United States." Two disclosures cut the other way: the reimbursement goal is to become "more in line with other NGS providers who have adopted similar strategies, such as FMI and Guardant," and competitors "may have longer operating histories; larger customer bases; greater brand recognition and market penetration." Competition names Foundation Medicine (Roche), Caris, Guardant, Natera, Neogenomics, Quest, LabCorp, Flatiron, IQVIA, ConcertAI and major CROs, all unranked. Scaled positions on both sides support parity, not a top position. | fast follower The FY2025 10-K itself places AMD behind the pace-setter in its two most important markets: 'Our principal competitor in the supply of discrete graphics is Nvidia, who is the discrete GPU market share leader,' and the risk factors state Nvidia 'leverages its market position in data center GPU, financial resources, and proprietary software ecosystem to promote its systems and influences customers who do business with us.' AMD is nonetheless executing 'an annual cadence of leadership for AMD Instinct solutions' and has won gigawatt-scale commitments (OpenAI, 6 GW) — following fast and closely, not setting the pace. |
| Pricing power | weak The 10-K's risk factors state 'Our competitors' products, services and technologies may be cheaper or provide better functionality or features than ours, which has resulted and may in the future result in lower-than-expected selling prices or demand for our products.' They also say long-term contract pricing 'is generally fixed or agreed at the time of contracting', so if market pricing for comparable capacity rises during a contract's term 'we will not benefit from those increases with respect to capacity already committed under our existing long-term contracts', and that if customers suffer a downturn or discontinue its services it 'may be compelled to offer more flexible terms, lower our prices or risk losing a significant customer.' The stored XBRL fundamentals from this 10-K show FY2026 revenue of about $707.0M against an operating loss of about $1,046.7M, so there is no margin record yet that evidences pricing power. | weak Diagnostics prices are administered: after a CPT code is set, "CMS, establish payment levels and coverage rules under Medicare," delegating to MACs, while "Private payers establish their rates and coverage rules independently." Average revenue per Oncology test rose from about $1,510 to about $1,600, "driven primarily by increased Medicare reimbursement rates" — the payer moved the price. Collection came on "approximately 55% of our clinical oncology NGS tests and 50% of our hereditary tests," and payer "cost-saving initiatives... are likely to result in pricing pressures." The moderate pocket is confined to Data and applications ($316.4M of $1,271.8M). | moderate FY2025 gross margin was 50%, up from 49% in 2024 per the 10-K — healthy and improving — but the same filing attributes average-selling-price pressure to competitors: Intel 'uses its microprocessor market position to price its products aggressively and target our customers and channel partners with special incentives. These aggressive activities have reduced and may reduce our unit sales and average selling prices for many of our products'. Pricing power is real but contested. |
| Summary | Per its FY2026 10-K, IREN is a vertically integrated AI Cloud Services platform that owns the data center, compute and software layers, underpinned by executed grid connection agreements, letters of agreement or equivalents representing ~5GW and a further multi-GW development pipeline. Its commercial record is concentrated: a five-year Microsoft agreement (~$9.7B total contract value, Horizon 1 delivered and accepted in August 2026, Horizons 2-4 targeted for delivery in phases in calendar Q4 2026) and a five-year ~$3.4B NVIDIA cloud contract together make up a substantial majority of contracted revenue. It holds NVIDIA Preferred Partner and Exemplar Cloud status (HGX B300 and GB300 NVL72) and a strategic partnership intended to support the deployment over time of up to 5GW of NVIDIA DSX-aligned infrastructure, which the filing cites among arrangements that are non-binding or subject to conditions, with no assurance as to the extent of deployments. Against that, the filing names hyperscaler and specialized competitors (AWS, Google Cloud, Azure, Oracle, CoreWeave, Nebius, Crusoe, Lambda, Nscale, SpaceX), many of which it says have greater resources, customers that can use or develop their own solutions, and possible delays to Texas energization from changes to ERCOT's Batch Zero procedures, so the power position is not a demonstrated advantage over rivals and its value depends on execution. | Tempus AI's durable asset is an installed provider network and the longitudinal data rights attached to it, not any single assay: connections to more than 55% of practicing U.S. oncologists and more than 65% of academic medical centers, 700+ data connections across 5,000+ institution sites. That network makes the linked clinical-molecular database sellable — Insights serves 19 of the 20 largest public pharmaceutical companies by 2024 revenue at roughly 126% net revenue retention, and the trial network (1,400+ trials signed, 40,000+ patients identified) depends on the sequencing base. Economics are improving: revenue up 83% to $1.27B, net loss down from $705.8M to $245.0M. It stops short of wide on the filing's own terms — a $2.4B accumulated deficit, payment on only ~55% of oncology and ~50% of hereditary tests, and Diagnostics prices set by CMS, MACs and private payers. | AMD's edge is architectural design IP executed on someone else's fabs. The FY2025 10-K describes a full-stack portfolio — EPYC server CPUs, Instinct AI accelerators, Ryzen client parts where 'AMD was the first company to integrate a dedicated neural processing unit (NPU) on the same SoC as an x86 CPU for AI PCs,' Radeon graphics, Pensando networking and Versal adaptive SoCs — backed by roughly 18,900 patent matters worldwide, and it discloses gigawatt-scale customer commitments (an October 2025 agreement with OpenAI 'to deploy 6 gigawatts of AMD GPUs,' first gigawatt on Instinct MI450). The same filing bounds that moat: Nvidia is 'the discrete GPU market share leader' and 'leverages its market position in data center GPU, financial resources, and proprietary software ecosystem'; Intel 'uses its microprocessor market position to price its products aggressively'; Arm architectures and customers who 'internally develop products to support similar AI workloads' are named as encroaching; and every wafer at 7 nm or below comes from TSMC. Revenue grew 34% in FY2025 with Data Center up 32%, but the filing's own competitive framing places AMD in the challenger seat rather than the entrenched one. |
| Chain position | Downstream AI-cloud operator: owns grid-connected data centers and deploys NVIDIA/AMD GPU systems it rents to hyperscalers, frontier labs, AI developers and enterprises (anchor customers Microsoft and NVIDIA). | Tempus sits between clinical care and pharma R&D. It ingests data from 5,000+ institution sites, runs five CAP/CLIA labs (Chicago, Atlanta, Raleigh, Aliso Viejo, Minneapolis via OneOme), and sells Diagnostics to providers and payers ($955.4M FY2025, +111%, administered) and Data and applications to pharma ($316.4M, +31%, negotiated). The clinical output is the data business's input. | A fabless merchant-silicon designer sitting one layer above the foundries: the FY2025 10-K states AMD relies on TSMC 'for the production of all wafers for microprocessor and GPU products at 7 nanometer (nm) or smaller nodes' and primarily on GLOBALFOUNDRIES above 7 nm, with packaging and test performed by Asia-Pacific ATMP partners — so its cost, supply and cadence are inherited from partners it does not control. |
| Products (share / barrier) |
|
|
|
| Long-horizon vote | -0.20 at weight 0.20 · swarm bearish Editorial prior, not backtested. | -0.01 at weight 0.20 · swarm neutral Editorial prior, not backtested. | +0.06 at weight 0.20 · swarm bullish Editorial prior, not backtested. |