Hyperscaler Capex Arms Race: Four Cloud Platforms Battling for AI Workload Dominance
The numbers are staggering and accelerating. Alphabet just guided $175–185 billion in 2026 capital expenditure. Amazon plans to spend roughly $200 billion, "predominantly in AWS." Microsoft's quarterly capex nearly doubled year-over-year to $29.9 billion in its December quarter. Oracle, the upstart challenger, is scaling from $2.8 billion to $12 billion in quarterly capex in barely a year. The hyperscaler capex arms race is the defining capital allocation story in technology — and the gap between winners and also-rans will compound for years.
Why This Theme Matters Now
AI workloads — training foundation models, running inference at scale, powering agentic applications — require compute infrastructure that didn't exist three years ago. The four major cloud platforms are in a land grab to build that infrastructure first, locking in customers through long-term commitments. AWS alone has a $200 billion backlog. Oracle's remaining performance obligations exploded to $523 billion. Microsoft's commercial RPO hit $400 billion. The companies spending the most now will own the AI compute layer for the next decade.
The Platforms: Who Is Winning the Buildout
We examined the four cloud platforms at the center of this spending cycle — from the incumbent leaders to the aggressive challenger — to assess where capital is being deployed most effectively and which platforms are translating capex into revenue growth.
1. Amazon (AMZN) — The Scale Leader Doubling Down
AWS is a $142 billion annualized run-rate business that reaccelerated to 24% growth in Q4 2025 — the fastest in thirteen quarters. Amazon's custom silicon business (Graviton, Trainium) has crossed $10 billion in annual revenue run rate, and Bedrock's customer spend is growing 60% quarter-over-quarter.
Amazon's Q4 2025 capex hit $39.5 billion, up 42% year-over-year, with management guiding to ~$200 billion for 2026. The sheer scale is unmatched: AWS plans to double capacity by 2027 and added 3.8 gigawatts of power in the past 12 months. The risk is FCF compression — trailing twelve-month free cash flow dropped to $7.7 billion as capex consumed nearly all operating cash flow.
| Metric | Value |
|---|---|
| Market Cap | $2,283B |
| Revenue (TTM) | $717B |
| Revenue Growth | 12% YoY |
| EBITDA Margin | 23% |
| P/E (fwd) | 27.7x |
| 1Y Price Return | +7% |
AWS's scale and backlog make it the default choice for enterprises, but the $200B capex commitment is a bet that AI demand will sustain for years.
2. Microsoft (MSFT) — The Enterprise AI Monetization Machine
Microsoft Cloud surpassed $50 billion in quarterly revenue for the first time in its December quarter, up 26% year-over-year. Azure is taking share, and the AI monetization flywheel — Copilot seats up 160% YoY, GitHub Copilot at 26 million users — is generating real revenue on top of infrastructure.
Quarterly capex reached $29.9 billion in the December quarter, nearly doubling from $15.8 billion a year earlier. Microsoft's advantage is that its AI spend translates into both infrastructure revenue (Azure) and application revenue (Copilot, Dynamics). Commercial RPO of $400 billion with a 2-year weighted average duration gives exceptional visibility.
| Metric | Value |
|---|---|
| Market Cap | $3,006B |
| Revenue (TTM) | $305B |
| Revenue Growth | 17% YoY |
| EBITDA Margin | 63% |
| P/E (fwd) | 23.0x |
| 1Y Price Return | +6% |
The highest-margin hyperscaler with the broadest AI monetization surface. Microsoft's ability to convert capex into application-layer revenue gives it the best return profile in the group.
3. Alphabet (GOOGL) — The Infrastructure Dark Horse
Google Cloud revenue surged 48% in Q4 2025, with AI revenue as a primary growth driver. Alphabet guided 2026 capex at $175–185 billion — a near-doubling from the ~$91 billion spent in 2025 — reflecting management's conviction that AI infrastructure demand is still supply-constrained.
Alphabet's differentiation lies in its vertically integrated AI stack: custom TPUs, the Gemini model family (serving costs down 78%), and 120,000+ enterprises using Gemini. Cloud backlog grew 46% quarter-over-quarter. Quarterly capex jumped to $27.9 billion in Q4 from $14.3 billion a year ago. Unlike peers, Alphabet still generated $73 billion in trailing twelve-month free cash flow — funding the buildout largely from search advertising cash flows.
| Metric | Value |
|---|---|
| Market Cap | $3,734B |
| Revenue (TTM) | $403B |
| Revenue Growth | 15% YoY |
| EBITDA Margin | 45% |
| P/E (fwd) | 27.1x |
| 1Y Price Return | +85% |
The best-positioned to self-fund the capex race. Google Cloud's 48% growth and custom silicon advantage make Alphabet the dark horse in the AI infrastructure race.
4. Oracle (ORCL) — The Challenger With the Biggest Backlog
Oracle's remaining performance obligations reached $523 billion, with cloud infrastructure revenue growing 66% year-over-year. The company has transformed from an enterprise database vendor into a credible hyperscaler, operating 147 live cloud regions with 64 more planned.
Oracle's capex trajectory is the steepest in the group: from $2.8 billion in Q3 FY2024 to $12 billion in Q2 FY2026 — a 4x increase in under two years. Management expects OCI revenue to grow 77% to $18 billion in FY2026 and projects a path to $144 billion. The cost: free cash flow has turned deeply negative at -$10 billion over the last two quarters, and the stock has pulled back 44% in the past month.
| Metric | Value |
|---|---|
| Market Cap | $469B |
| Revenue (TTM) | $61B |
| Revenue Growth | 11% YoY |
| EBITDA Margin | 44% |
| P/E (fwd) | 21.7x |
| 1Y Price Return | +8% |
The cheapest valuation and fastest cloud infrastructure growth, but the highest execution risk. Oracle's massive backlog must convert to revenue before the balance sheet strains further.
5. Dell Technologies (DELL) — The Picks-and-Shovels Beneficiary
Dell's Infrastructure Solutions Group is the direct hardware beneficiary of the hyperscaler buildout, with total revenue growing 19% year-over-year. Every dollar of cloud capex flows through server, storage, and networking vendors — and Dell is the largest enterprise server provider.
Dell's Q4 FY2026 revenue hit $33.4 billion, with operating income of $3.1 billion. The AI server pipeline continues to expand as all four cloud platforms scale their GPU clusters. Dell trades at a significant discount to the hyperscalers despite being a direct beneficiary of their spending.
| Metric | Value |
|---|---|
| Market Cap | $99B |
| Revenue (TTM) | $114B |
| Revenue Growth | 19% YoY |
| EBITDA Margin | 7% |
| P/E (fwd) | 12.8x |
| 1Y Price Return | +56% |
The cheapest way to play the capex arms race without betting on which cloud platform wins. Dell benefits regardless of who captures the AI workloads.
The Verdict: Ranking the Picks
Microsoft offers the best risk-adjusted exposure: highest margins, broadest AI monetization (infrastructure + applications), and reasonable valuation at 23x forward earnings. Alphabet is the strongest self-funded grower — 48% cloud growth, custom silicon, and $73 billion in annual FCF make the $175–185 billion capex guide credible. Amazon has the largest scale and backlog but the thinnest FCF cushion at current capex levels. Dell is the value play — at 12.8x forward earnings, it captures upside from all hyperscalers' spending without platform risk. Oracle has the most aggressive growth trajectory and cheapest cloud valuation, but the 44% drawdown reflects real concerns about FCF burn and execution on its $523 billion backlog.
Risks to Watch
- AI demand plateau: If enterprise AI adoption slows, the massive capex commitments become stranded assets with accelerating depreciation
- Margin compression: All four platforms are guiding depreciation acceleration in 2026; operating margins will face pressure before revenue scales
- Geopolitical supply chain risk: GPU supply constraints (NVIDIA allocation), power grid limitations, and data sovereignty regulations could slow buildout timelines
What to Monitor
- Cloud revenue growth rates vs. capex growth rates: The gap between spending and revenue conversion is the key health metric — watch for deceleration in cloud bookings while capex continues rising
- Free cash flow inflection: Oracle and Amazon need FCF to stabilize; if it deteriorates further, capital markets may reprice the stocks