At the beginning of June 2026, two leading AI cloud giants sent rare, divergent signals—Amazon closed at $261.26, down 3.47%, while Microsoft closed at $460.52, up 2.28%. Around the same time, Morgan Stanley projected global AI infrastructure capital expenditures to reach $800 billion, rising to $1.1 trillion by 2027.
Complete Data for June 2 (USD)
| Item | Amazon (AMZN) | Microsoft (MSFT) |
|---|---|---|
| Closing Price | 261.26 | 460.52 |
| Daily Change | -3.47% | +2.28% |
| 52-Week High | 278.56 | 555.45 |
| Off High | -4.57% | -16.98% |
| Market Cap | ~$2.84T | ~$3.15T |
Data Source: Nasdaq Market Overview, June 2, 2026; FinanceCharts
Comparing Revenue Fundamentals: Key Signals from Q1 2026 Earnings
In the Q1 2026 earnings season (Microsoft FY2026 Q3), both companies delivered results that exceeded expectations, driven by their AI businesses.
Amazon’s Q1 report showed total net sales of $181.5 billion, up 17% year-over-year. AWS revenue reached $37.6 billion, a 28% increase year-over-year, beating market expectations of $36.6 billion and marking the fastest growth in nearly four years. AWS operating income was $14.2 billion, with an operating margin of 37.8%, significantly higher than the company’s overall margin. Management’s Q2 guidance projected net sales between $194 billion and $199 billion (implying 16%-19% YoY growth) and operating income between $20 billion and $24 billion.
Microsoft reported revenue of $82.886 billion for the same period, up 18% year-over-year and about $1.5 billion above market expectations. The Intelligent Cloud segment generated $34.7 billion in revenue, up 29% YoY; within this, Azure and other cloud services grew 40% YoY, surpassing the expected 37% and up from 39% the previous quarter. Annualized AI revenue (ARR) exceeded $37 billion, up 123% YoY. Management expects Azure to see a "modest acceleration" in growth in the second half of the year.
In terms of total revenue, AWS’s cloud business ($37.6 billion) still leads Microsoft’s Intelligent Cloud segment ($34.7 billion), but the gap is narrowing. Looking at growth rates, Azure’s 40% surge far outpaces AWS’s 28%, and the gap is widening—last quarter, Azure grew 39% while AWS was around 20%.
However, two important clarifications are needed when comparing these growth rates:
Differences in Reporting Scope. Microsoft’s Intelligent Cloud segment includes Azure, SQL Server, Windows Server, and other enterprise products. Azure’s own growth rate exceeds the segment’s 40% figure and aligns more closely with independent estimates, which put Azure’s growth at about 31%. AWS’s 28% growth reflects pure IaaS/PaaS revenue.
Strength of Growth Signals. Azure’s 40% growth came after breaking a two-quarter slowdown, providing a clear directional signal. While AWS’s 28% growth is lower, it still marks a four-year high within its own cycle.
Market Share: Static Readings and Incremental Shifts
In terms of global cloud infrastructure spending, AWS’s dominant position remained unshaken in Q1. According to Synergy Research Group’s Q1 2026 data, AWS accounts for about 28%-30% of global cloud infrastructure spend, Microsoft holds 21%-25%, and Google Cloud 13%-14%. Together, the three control roughly 67% of the global public cloud market.
However, these static market share figures mask structural changes in incremental growth. The same Synergy Research data shows AWS revenue grew about 19% YoY, Azure about 31% (with roughly 12 percentage points attributed to AI services), and Google Cloud about 63%. Other public estimates put Azure’s growth around 30% and Google’s around 40%, reshuffling the growth rankings: Google leads, Azure is in the middle, and AWS lags.
In the current AI-driven cloud growth landscape, Azure is closing the absolute revenue gap with AWS at a pace well above the industry average.
Divergence in Profit Quality: The "Structural Dividend" and "Cost Pressure" of AI Revenue
If revenue growth sets the market’s pricing direction for cloud businesses, profit structure determines the sustainability of that pricing. AWS and Azure show distinct characteristics in their AI revenue share and profit structures.
AWS’s annualized AI services run rate has surpassed $15 billion, maintaining triple-digit growth year-over-year. The number of tokens processed by the Bedrock platform in Q1 exceeded the total from all previous years, with customer spending up 170% quarter-over-quarter. Amazon’s in-house chips (Graviton, Trainium, Nitro) generated over $20 billion in annualized revenue, also growing at triple digits. Anthropic saw a net new annual recurring revenue of about $21 billion in the prior quarter.
AWS’s high profit structure comes from its hybrid "model distribution + compute rental" approach, with Bedrock generating higher marginal profits per revenue unit than traditional AI IaaS models. SemiAnalysis estimates Bedrock’s EBIT margin at around 55%, giving AWS a relative profit advantage even in a high capex environment.
Microsoft’s annualized AI revenue exceeded $37 billion, up 123% YoY. Copilot added 5 million paid seats in a single quarter, and the commercial version of Office 365 Copilot has sold over 20 million seats. However, declining gross margins are a concern—this quarter’s gross margin was 67.6%, the lowest since 2022. The Intelligent Cloud segment’s operating margin was 39.5%, down about 1.8 percentage points YoY, as depreciation costs rose with data center expansion.
The core difference between these paths: Azure prioritizes expanding AI revenue scale, leveraging Copilot’s deep integration with enterprise software to drive broader commercial adoption—but it also bears higher depreciation and compute costs. AWS’s absolute AI revenue is about 40% that of Microsoft, but through Bedrock’s margin leverage and in-house chip cost optimization, it has carved out a differentiated advantage in profit quality.
Capital Expenditures: Two Sides of the Same Coin—Scale Expansion and Margin Pressure
In the AI cloud competition narrative, capital expenditure is the inescapable variable. Morgan Stanley’s Andrew Sheets projects that US tech giants’ AI-related capex will reach about $800 billion in 2026—nearly double 2025 estimates and triple 2024’s. By 2027, it’s expected to rise to about $1.1 trillion.
From a competitive logic perspective, capex and profit are naturally inversely related—scale expansion means short-term profitability gives way to long-term capacity building, as seen in Q1 earnings where aggressive capex plans weighed on guidance.
Amazon expects full-year 2026 capex of about $200 billion, primarily for AI infrastructure. Q1 capex already hit $43.2 billion, and free cash flow shrank sharply from $25.9 billion a year ago to $1.2 billion, driven by a $59.3 billion YoY increase in equipment spending. AWS accounts for just 20.7% of Amazon’s total revenue but contributes roughly 60% of operating profit.
Microsoft projects FY2026 capex at about $190 billion, up 61% YoY, with about $25 billion attributed to rising component costs (especially high-bandwidth memory). This quarter’s capex was $31.9 billion, below market expectations of $34.9 billion.
While both companies’ capex levels are similar, their profit structures mean they absorb high capex differently: AWS’s high margins better offset capex pressure, but it still faces a sharp rise in free cash flow constraints. Microsoft must continually balance Azure’s rapid growth against margin compression.
Strategic Divergence: Compute Efficiency vs. Enterprise Adoption
In May 2026, Microsoft announced a revised partnership with OpenAI: Microsoft will no longer share revenue with OpenAI, instead, OpenAI will pay Microsoft for services; in exchange, Microsoft relinquished exclusive rights to OpenAI’s models and technology. Previously, Azure held multi-year exclusive deployment rights for OpenAI models—a core moat in AI cloud competition. Under the new terms, Microsoft retains IP usage rights through 2032 but loses exclusive deployment advantages.
This shift in pricing power marks a new phase in AI cloud competition—rivalry is now spreading across compute cost competitiveness, model ecosystem richness, and enterprise-grade commercial integration.
From a compute cost perspective, Amazon’s in-house chip business (Trainium, Graviton, Nitro) now generates over $20 billion in annualized revenue. Bedrock’s margin leverage is forming a structural alternative to GPU dependency, allowing AWS to sustain triple-digit AI revenue growth without significantly sacrificing margins. From the enterprise adoption angle, Microsoft’s deep integration of Copilot with Office 365, Teams, LinkedIn, and other core enterprise software has created a differentiated edge in AI monetization speed and market penetration—Copilot’s weekly active users now match Outlook’s. This is a contest on two fronts: short-term, it’s about enterprise rollout speed and revenue scale; long-term, it’s about compute cost structure and profit quality.
From AI Cloud Trends to Asset Allocation: How Gate Is Participating
The competitive landscape between AMZN and MSFT in AI cloud is not just a core narrative for the cloud and tech sectors—it also directly impacts crypto investors’ asset allocation decisions.
On June 1, 2026, Gate officially launched US stock and ETF trading services. Users can directly buy, hold, and sell US stocks and ETFs using USDT—no need to convert crypto to fiat or transfer funds to another brokerage account. As of June 3, 2026, Gate supports over 10,000 stocks and ETF assets, covering major US exchanges such as NYSE, Nasdaq, NYSE Arca, NYSE American, and BATS.
Both AMZN and MSFT are now listed for trading on Gate. Users can trade these stocks directly using their USDT balance within the Gate platform. This marks the latest step in Gate’s multi-phase TradFi expansion strategy: launching TradFi CFDs in January 2026, expanding to tokenized stocks and ETFs (24/7 trading) in March, and rolling out real US stock and ETF spot trading in June—offering physical ownership, SIPC-member broker custody, and USDT settlement.
Key Points for Participating in US Stock Allocation via Gate:
Asset Authenticity: Stocks purchased on Gate—such as AMZN and MSFT—are real underlying assets traded in sync with Nasdaq, not tokenized derivatives or CFD wrappers. The stocks are held in custody by a compliant broker-dealer licensed in the US, a SIPC member, ensuring investor protection for securities assets under applicable conditions.
Deposit and Withdrawal Process: Users do not need to convert USDT to fiat or transfer funds to another brokerage. They can buy and sell US stocks directly using their USDT balance in the Gate wallet. This extends USDT’s use case from digital asset trading to global stock allocation, allowing crypto holdings and stock investments to coexist within the same ecosystem.
Trading Hours and Costs: Gate’s US stock spot trading operates during standard US market hours and supports intraday trading. The tokenized stock section, launched in March 2026, offers 24/7 trading to meet various investment timeframes. Compared to tokenized stocks and CFD contracts, spot stock trading incurs zero holding costs, making it better suited for long-term allocation.
Account and Leverage Options: The unified account interface lets users track positions, P&L, trade history, and corporate actions like dividends. For those seeking higher volatility, Gate also offers MSFT perpetual contracts (USDT-settled) with 1x to 20x long and short leverage. Users can freely switch between spot, tokenized assets, and derivatives within the same account framework, according to their risk preferences and investment goals.
Conclusion: The Logic of Market Pricing Is Being Rewritten
The divergence in AMZN and MSFT stock prices in early June 2026 is a structural signal that needs to be understood in a broader, long-term context. Within the same AI infrastructure investment narrative, the market is applying distinctly different pricing logics to AWS and Azure.
Fundamentally, both companies delivered Q1 results that beat expectations, both face ongoing margin pressure from high capex, and both achieved triple-digit annualized AI revenue growth. Yet, in the market’s valuation models for AI cloud businesses, the slope of growth, quality of profits, and sustainability of AI strategy are all being reassessed.
AWS’s 28% growth outpaces its historical average and it still leads in market share, but it faces structural constraints from a lower AI revenue share and slower growth compared to key rivals. At the same time, the margin leverage from the Bedrock platform is shifting perceptions of its profitability.
Azure’s 40% growth highlights strong AI-driven momentum, and Copilot’s commercial progress provides tangible enterprise demand for the AI narrative. However, capex-driven margin compression and the loosening of exclusive OpenAI partnership terms present ongoing valuation challenges.
Morgan Stanley’s forecast of $800 billion to $1.1 trillion in capex for 2026 and 2027 signals that the expansion cycle for AI infrastructure is far from over. In this cycle, hyperscale cloud providers are moving beyond simple market share battles to a multidimensional contest involving compute costs, model ecosystems, and profit structures. For market participants, understanding the valuation differences between AMZN and MSFT ultimately boils down to a fundamental question: In this wave of AI infrastructure investment, is the anchor for valuation the momentum of revenue growth or the sustainability of profit quality?
