超大規模市場追蹤（2026年第一季）：資本支出創歷史新高，自由現金流依然低迷，但樂觀情緒依舊不變

2026年第一季概覽

2026年第一季，超大規模業務營收達7,897億美元（單以季度為基礎增17.4%），年化營收達3.136兆美元（年增14.9%）。年化資本支出飆升至5,921億美元（年增70.4%），資本密集度達到營收的18.9%，遠高於電信業約16%的水準。有形固定資產淨成長51.9%。淨利率達到創紀錄的22.5%，但自由現金流利潤率降至11.3%，為該資料庫自2011年開始收集數據以來的最低水準。 2026年第一季員工人數較去年同期成長6.3%。這份完整報告分析了這些數據背後的因素，哪些公司表現良好，哪些公司舉步維艱，以及這些數據對未來市場趨勢的啟示。

閱讀這份報告的理由

MTN Consulting自2017年第四季起，透過詳盡的季度分析持續追蹤該市場，資料庫可追溯至2011年第四季。該公司在該市場擁有長期深入的研究經驗，掌握著專有工具和關於全球領先超大規模資料中心業者（原稱刮器）的一致且可比的數據，涵蓋收入、資本支出、研發費用、盈利、員工人數和資產負債表指標等。其涵蓋的公司超過12家，包括亞馬遜、Alphabet、微軟、Meta、蘋果、甲骨文、阿里巴巴和騰訊，以及Coreweave、Nebius、快手和小米等新興參與企業。沒有其他分析公司能夠提供如此深入且全面的市場分析。

該公司一年半以來一直認為該市場存在泡沫，這一觀點貫穿整個分析。 2026 年第一季報告包含詳細的公司層級細分資料、區域資料、供應商層級的銷售基準，以及 2026 年至 2028 年資本支出前景的三種情境預測。對於任何計劃根據超大規模公司的支出趨勢制定投資、採購或策略決策的人士而言，本報告都是理想之選。

注記：

• （1）本報告是 MTN Consulting 對超大規模市場的第 34 份季度評估報告，也是自 2017 年第四季開始的一系列報告的一部分。該公司的數據和分析涵蓋了 61 個季度，從 2011 年第一季到 2026 年第一季。
• （2）本次調查中包含的公司包括幾家最近新增的公司：CoreWeave（2024 年新增）、快手、Nebius（Yandex 的衍生公司）和小米（2025 年第三季新增）。
• （3）在 2025 年第四季版中，該行業的名稱從「Webscale」更改為更通用的「Hyperscale」。

目錄

• 1. 報告要點
• 2. 展望
• 3. 分析
• 4. 主要統計數據
• 5. 公司詳細分析
• 6. 企業標竿管理
• 7. 按地區分類
• 8. 原始數據
• 9.外匯
• 10. 關於我們

Snapshot view of 1Q26

In 1Q26, hyperscale revenues reached $789.7B (single-quarter, +17.4% YoY), with annualized revenues reaching $3.136T (+14.9% YoY). Annualized capex surged to $592.1B (+70.4% YoY), driving capital intensity to 18.9% of revenues, well above the telecom sector’s ~16% level. Net PP&E grew +51.9% YoY. Net profit margin reached a fresh record of 22.5%, but FCF margin fell to 11.3%, its lowest since our database begins in 2011. Headcount grew 6.3% YoY in 1Q26. The full report analyzes what is driving each of these numbers, which companies are winning and losing, and what the data says about where this market goes next.

Why read this report

MTN Consulting has been tracking this market with quarterly deep dives since 4Q17, and our database begins in 4Q11. We have a long history of in-depth coverage of this market, and proprietary tools and consistent, comparable data across revenues, capex, R&D, profitability, employment, and balance sheet metrics for the world’s leading hyperscalers (previously called “webscalers”). We cover Amazon, Alphabet, Microsoft, Meta, Apple, Oracle, Alibaba, Tencent, and more than a dozen other players, including newer entrants like Coreweave, Nebius, Kuaishou, and Xiaomi. No other analyst firm covers this market with this depth or our data coverage.

We have been calling this market a bubble for over a year and a half, and that view is reflected throughout the analysis. The 1Q26 report includes detailed company-by-company breakouts, regional data, vendor revenue benchmarks, and three capex outlook scenarios for 2026-2028. If you are making investment, procurement, or strategic decisions that depend on where hyperscale spending goes from here, this report is built for you.

Notes:

• (1) This is MTN Consulting’s 34th quarterly assessment of the hyperscale market, part of a series we launched in 4Q17; our data and analysis spans the 1Q11-1Q26 timeframe, i.e. 61 quarters.
• (2) The companies in our study include several recent additions: CoreWeave (added in 2024), and Kuaishou, Nebius (Yandex spinoff), and Xiaomi (added in 3Q25).
• (3) The 4Q25 edition adopts a new term for the sector, swapping “webscale” with the more common “hyperscale”.

Revenue: Topline reaches $789.7B (+17.4% YoY), led by Meta (+33%), Alphabet (+22%), and Coreweave (+116%)

Hyperscale revenues reached $789.7B in 1Q26, up 17.4% YoY, with annualized revenues rising to $3.136T (+14.9% YoY). Coreweave (+116% YoY) and Nebius (+684%) are the fastest-growing companies in our coverage, but remain outliers in absolute scale and geographic reach. Among established players, Meta (+33%) continued its AI-driven advertising renaissance, with AI-powered ranking, targeting, and recommendation systems driving higher CPMs across Facebook and Instagram. Alphabet (+21.8%) was led by GCP acceleration, YouTube, and early Gemini enterprise adoption. HPE (+40%, absorbing Juniper Networks into its networking business) and Microsoft (+18.3%) also outperformed the sector average. Amazon (+16.6%) was in line with the market despite strong AWS momentum. At the other end: Xiaomi revenues declined (-6.4%), while Baidu (+3.8%) and Alibaba (+8.1%) continued to underperform, reflecting subdued Chinese macro conditions, e-commerce price competition from JD and Pinduoduo/Temu, and Baidu’s weakness in its legacy advertising business. Soaring memory prices and a Chinese government shift away from IoT subsidies impacted Xiaomi negatively.

Advertising remains central for several firms. Meta is the most exposed, with ads still driving nearly all revenue, despite efforts to diversify into hardware and AI platforms. Alphabet’s non-ad share has risen above 25%. Amazon is approaching 10% of revenues from ads. Ad-dependent companies face heightened risk given concerns about US consumer spending in 2026, which have grown more pronounced following the energy market disruption created by the Trump administration’s actions in Iran. A further open question is whether scaled AI platforms will rely heavily on ads, given slow traction for paid subscription models outside early adopters.

Capex: Annualized spend hits record $592B (+70% YoY) as AI spending race outpaces economic rationality

Annualized capex reached $592.1B in 1Q26, up 70.4% YoY, with capital intensity rising to 18.9% of revenues. Consensus 2026 capex estimates have witnessed upward revisions as every company cites committed backlog, unprecedented demand, and long-term return confidence. Some of this is real: the combined contracted backlog of Microsoft, Google Cloud, AWS, and Oracle surpassed $2 trillion in 1Q26, reflecting massive committed demand for AI and cloud services, though that $2T includes a good amount of circular financing.

Moreover, the capex escalation carries the character of an arms race, every company afraid to signal less AI commitment than its peers, and a sunk-cost logic: once hundreds of billions are committed, stopping feels expensive than continuing. Stopping means writing down assets already on the books, telling investors the strategy failed, and ceding ground to rivals who kept building. So, companies keep spending. The capex cycle is partly bluster, partly competitive fear, and partly a genuine attempt to reshape markets before a reckoning catches up.

Power and connectivity continue to be crucial constraints shaping the pace of AI buildouts. In the last few months, both Meta (February) and Amazon (June) have signed multi-billion US dollar contracts for fiber with Corning. Since it’s flush with cash and wants to keep the AI party going, NVIDIA is helping Corning ramp up production: in May, NVIDIA invested $500M in Corning to expand US optical manufacturing tenfold and increase US fiber production capacity by over 50%. These are not routine procurement deals, but signals that fiber is a guiding constraint on hyperscale expansion.

Power is playing a similar role - that of a strategic bottleneck, not just an operating cost line item; there are many examples of this. Google has committed $20B to new clean energy projects specifically to power future data centers. Private equity player KKR recently launched Helix Digital Infrastructure, a new platform company focused on developing hyperscale data centers with pre-secured power supply. Microsoft was forced to abandon a planned data center build in Kenya recently due to insufficient power capacity in country. As the energy constraint becomes a primary limiter, dollars that once flowed to GPU vendors are now flowing to utilities and energy developers.

Custom silicon: Amazon Trainium has reached a $20B annual run rate as it looks to cut NVIDIA’s dependency. Amazon’s Trainium-based chip business is running at a $20B annual revenue rate (growing triple digits). CEO Andy Jassy stated it would generate $50B annually if sold externally like a conventional chip company. Trainium2 is sold out; Trainium3, which began shipping in early 2026, is nearly fully subscribed. Amazon estimates Trainium will save it tens of billions in annual capex relative to continued NVIDIA reliance and provide several hundred basis points of margin advantage. Google’s TPU program and Meta’s custom Broadcom-developed silicon are on similar trajectories. As custom silicon matures, it will reduce NVIDIA’s share of Network/IT capex.

R&D, and Net PP&E: capex intensity remains above R&D. R&D intensity settled at 13.0% of revenues in 1Q26, sustaining its position below capex intensity (18.9%). For most of this sector’s history, R&D tracked above capex as a share of revenues, reflecting a software-centric innovation model. That reversal deepened in 1Q26. The sector has shifted from a code-first to a hardware-first growth model. This is not permanent; as custom silicon displaces commercial GPU purchases and the infrastructure build matures, more capital will flow back toward software, model development, and monetization infrastructure. But in the near term the hardware land grab dominates. Net PP&E per employee is rising sharply as net PP&E grew +51.9% YoY against only 6.3% headcount growth. Large AI model builders such as Coreweave and Nebius are outliers.

M&A: M&A activity remains muted relative to capex: acquisitions take 12+ months to close, too slow for an arms race moving at this speed. Capital flows instead to infrastructure capex, targeted IP purchases, and acqui-hires. The Meta-Scale AI deal (~$14B+ for a 49% stake) illustrates the pattern: immediate access to talent and data-labeling IP without integration lag or regulatory friction.

China: Huawei Ascend 910D benchmarks near H100, DeepSeek runs on domestic silicon; the chip embargo is failing

Huawei’s Ascend 910D, expected in mid-2026, is being benchmarked as potentially matching or exceeding NVIDIA H100 performance. Huawei plans to deliver roughly 600,000 units of Ascend 910C in 2026, approximately double the prior year. DeepSeek’s V4-Pro model is trained on Huawei silicon. Baidu is running a 30,000-chip training cluster entirely on domestic processors and is preparing to spin off its chip arm (Kunlunxin) via a dual IPO in Hong Kong and Shanghai. The Stanford 2026 AI Index states that Chinese AI labs have effectively closed the performance gap with US counterparts. Chip embargoes have not delivered their intended outcome.

This matters for capex forecasts in two ways. First, if Chinese models are competitive at far lower investment levels, this directly undermines the narrative that massive US hyperscaler capex is the only viable path to advanced AI. Second, a high-profile Chinese breakthrough could trigger rapid reassessment of US spending plans, particularly if it coincides with weak macro data or disappointing AI revenue monetization.

Capex Outlook: Three scenarios for 2026-2028 - macro, AI monetization, and China determine which path

Capex in 2025 ended at just over $500B. Three potential scenarios shape the outlook. The high case (~$892B in 2026, ~$920B in 2027, ~$883B in 2028) follows current official guidance from Amazon ($200B), Alphabet ($180-190B), Microsoft (~$190B), Meta ($125-145B), and others. Some would call this the ‘base case’, but we do not believe current plans are sustainable at this pace. It assumes NVIDIA maintains pricing power, all announced projects proceed on schedule, and no macro or geopolitical event forces a recalibration.

The base case (~$683B in 2026, ~$678B in 2027, ~$654B in 2028) assumes more conservative execution: NVIDIA begins to lose share to custom silicon; some projects are consolidated or delayed; energy and power constraints create real deployment bottlenecks. This is our central view.

The low case (~$604B in 2026, declining to ~$498B in 2027 and ~$457B in 2028) assumes a market correction begins in mid-2026. The likely trigger is a combination of weak macro data, soft AI revenue monetization, and a high-profile Chinese AI breakthrough that resets global expectations. A delayed or canceled OpenAI IPO would be an early signal.

Profitability: Net margin hits record high while free cash flow margin stays at record low

The divergence between net margin and FCF margin has widened significantly. Net profitability reached a record 22.5% in 1Q26, the highest in our database going back to 2011. At the same time, FCF margin sits at 11.3%, the lowest on record. FCF is cash from operations minus capex and is generally a more reliable profit metric than net income. The gap reflects the brutal math of the AI arms race: because FCF is operating cash minus capex, the hyperscaler heavyweights risk cannibalizing their own liquidity with their immense capex spend. Microsoft, Alphabet, and Meta are reporting good net income results, but are eating into cash reserves as they race to build more GPU clusters. Amazon and Oracle remain in the same cycle, trading liquid cash for hard infrastructure with long payoff horizons.

On a per-employee basis, Apple leads at $773K FCF per employee (1Q26 annualized), followed by Meta at $619K, Alphabet at $331K, and Microsoft at $323K. Apple’s high figure reflects its deliberate decision to stay out of the GPU infrastructure arms race. By FCF margin: Tencent (31.8%) and Apple (28.6%) lead the sector, followed by Microsoft (22.9%) and Meta (22.4%). At the other end, Oracle (-20.7%) is in a heavy build-ahead phase where GPU capex hits immediately while revenues lag. Coreweave (-170.5%) and Nebius (-358.9%) are deeply negative as early-stage neoclouds scaling rapidly against contracted backlogs. Alibaba (-4.9%) and Baidu (-7.3%) are both negative, reflecting China’s AI ramp-up costs against weaker core revenues.

Regulatory fines and civil lawsuits represent a persistent, though minor, risk to profitability. Hyperscalers consistently treat this as a cost of doing business, often ignoring rulings, aggressively fighting them in court, and using public relations to minimize backlash, moving far from the earlier “don’t be evil” philosophy.

Employment: Headcount grows +6.3% YoY

Headcount grew +6.3% YoY in 1Q26, coming in just below the 2019-25 CAGR of 7.9%. Total headcount is a tricky metric in hyperscale, as the sector’s employee base is influenced heavily by logistics, fulfillment, and delivery employees at companies like Amazon, Alibaba, and JD.Com. For instance, JD.Com’s headcount rose by over 30% between 1Q25 and 1Q26, to over 800K.Among the more tech-centric hyperscalers, the direction is clearly downward.

The hyperscale business model is built around massive economies of scale, with investment concentrated in areas where the marginal cost of production can approach zero. Many hyperscaler executives would be happy to see headcount fall significantly and rely on AI platforms to run more of their operations over time. That is clearly the industry’s direction. Revenue per employee and net PP&E per employee have both made sizable gains in the last 2-3 years, and those trends are likely to continue.

Regional Revenues: Sovereignity and AI mandates driving Europe and MEA surge

Americas revenues reached $359.8B in 1Q26 (single-quarter, 45.6% of total, +16.0% YoY), below the sector average (+17.4% YoY). Asia Pacific contributed $258.7B (32.8%, +16.8% YoY). Europe reached $145.7B (18.4%, +21.8% YoY) and MEA $25.6B (3.2%, +21.0% YoY). Europe and MEA are growing fastest, a pattern that continued from 4Q25, reflecting ongoing international expansion by US-based hyperscalers into underpenetrated markets, partly accelerated by data sovereignty regulations that are pushing local cloud buildout. Asia Pacific’s more modest growth reflects Chinese macro headwinds and competitive pressures on Alibaba, Baidu, and JD. Europe’s structural outperformance reflects sustained demand from regulated industries, AI Act compliance-driven cloud migration, and sovereign AI initiatives.

Table of Contents

• 1. Report highlights
• 2. Outlook
• 3. Analysis
• 4. Key Stats
• 5. Company Drilldown
• 6. Company Benchmarking
• 7. Regional Breakouts
• 8. Raw Data
• 9. Exchange Rates
• 10. About

Figures and Charts

• 1. Capex outlook by scenario ($B)
• 2. Key Metrics: Growth rates, Annualized 1Q26/1Q25 vs. 2021-25
• 3. Hyperscale Revenues: Single-quarter & annualized (US$M)
• 4. Key Hyperscaler revenues: YoY % revenue growth in 1Q26
• 5. Annualized profitability: HYPERSCALERS
• 6. Free cash flow per employee, 1Q26 annualized (US$)
• 7. Key Hyperscaler free cash flow margins: 1Q26 annualized
• 8. Advertising revenues as % total (FY2025)
• 9. Annualized capex and R&D spending: HYPERSCALERS (% revenues)
• 10. Hyperscale capex by type, Annualized: 1Q15-1Q26 (US$M)
• 11. Network & IT capex as share of revenues, 1Q26 annualized
• 12. Key Hyperscaler R&D expenses, % revenues: 1Q26 annualized
• 13. Acquisition spending vs. capex spending, annualized (US$M)
• 14. Net PP&E per employee (US$’ 000) - 1Q26
• 15. Ranking the Hyperscale Network Operators: Revenues; R&D; Capex; Network & IT capex - 2025 & 1Q26 (US$B)
• 16. Annualized spending for key Hyperscalers since 2011 Capex: Network, IT and software
• 17. Share of Hyperscale spending by company, 1Q26 and 1Q25 annualized (Capex: Network, IT and software)
• 18. Energy consumption vs. Net PP&E for key Hyperscalers in 2024
• 19. Hyperscale vs. Telco Market: Annualized Capex (US$B)
• 20. Hyperscale vs. Telco Market: Annualized capital intensity
• 21. USA: Hyperscale capex total ($M) and % of global market, 2011-25
• 22. China’s Hyperscalers versus the big 4: Capex in 2025 ($B)
• 23. Chinese Hyperscale capex on the rise again ($M)
• 24. Hyperscale tech capex vs. DC-related chip revenues (annualized $M)
• 25. Hyperscale tech capex vs. Data center-related systems revenues (annualized $M)
• 26. Revenues: annual, single-quarter, and annualized (US$M)
• 27. Profitability (Net Profit; Cash from operations; Free cash flow): annual, single-quarter, and annualized (US$M)
• 28. Spending (R&D; M&A; Capex; Network & IT capex; Lease): annual, single-quarter, and annualized (US$M)
• 29. Cash & Short-term Investments: annual and single-quarter (US$M)
• 30. Debt (Total debt; Net debt): annual and single-quarter (US$M)
• 31. Property, Plant & Equipment: annual and single-quarter (US$M)
• 32. Key Ratios: Net margin; R&D/revenues; Capex/revenues; Network & IT capex/revenues; Free cash flow/revenues; Lease costs/revenues - annual and annualized (%)
• 33. Total employees
• 34. Revenue per employee, annualized (US$K)
• 35. FCF per employee, annualized (US$K)
• 36. Net PP&E per employee, annualized (US$K)
• 37. Revenues & Spending (US$M)
• 38. Revenues (US$M) & YoY revenue growth (%), single-quarter: by company
• 39. Revenues, annualized (US$M): by company
• 40. Annualized profitability margins: by company
• 41. Annualized capex and capital intensity: by company
• 42. Annualized capex and R&D spending as % of revenues: by company
• 43. Share of hyperscale network & IT capex, Annualized: by company
• 44. Total employees: by company
• 45. Annualized per-employee metrics (US$000s): by company
• 46. Net debt (debt minus cash & stock) (US$M): by company
• 47. Top 10 Hyperscale employers in 1Q26: Global market
• 48. Headcount changes in 1Q26 (YoY %): Global market
• 49. Net PP&E: USA vs. RoW (by company)
• 50. Net PP&E: total in $M and % global Hyperscale market (by company)
• 51. Energy consumption, MWh and % Hyperscale total (by company)
• 52. Share of Hyperscale energy consumption, net PP&E, and capex (by company)
• 53. Energy intensity relative to Hyperscale average and select data center-focused CNNOs (by company)
• 54. Energy intensity in Hyperscale sector, 2024: MWh consumed per $M in revenue
• 55. Capex/revenues (annualized): Company vs. Hyperscale average
• 56. Revenue per employee (US$000s) (annualized): Company vs. Hyperscale average
• 57. 2018 vs. 2025: company benchmark by KPI (Revenues, R&D, Net profit, Cash from operations, Capex, Free cash flow, Cash & short-term investments, Net PP&E, Total debt)
• 58. 2018 vs. 2025: company benchmark by key ratio (Capex/revenues; R&D/revenues; Net margin; FCF margin)
• 59. Top 8 Hyperscaler share vs. Rest of the market: by KPI (Revenues, R&D, Net profit, Cash from operations, Capex, Free cash flow, Cash & short-term investments, Net PP&E, Total debt)
• 60. Top 8 Hyperscalers benchmarking by Key ratio: Capex/revenues; R&D/revenues; Net margin; FCF margin)
• 61. Total Hyperscaler Market Revenues, by region: Latest CY; Latest Quarter; Annual trend (2011-25); Single quarter (1Q17-1Q26)
• 62. Hyperscaler Market: Revenues, single-quarter (YoY % change)
• 63. Regional revenues by operator: Latest CY; Latest Quarter; Annual trend (2011-25); Single quarter (1Q17-1Q26)
• 64. Top 10 operators by region: Latest CY; Latest Quarter