Meta Platforms (META.US) shares closed up 8.81% on July 1 after foreign media reported that Meta is building a cloud computing business and plans to sell AI computing power to external customers. The news triggered a chain sell-off in AI hardware stocks. Meta has committed to capital expenditures of $125 billion to $145 billion in 2026.
Closing Stock Reaction: META Up Over 8%, Six AI Hardware Stocks Close Lower
After the announcement on July 1, 2026, market reaction showed clear sector divergence: the software sector posted its second-largest single-day excess return over the semiconductor sector in the past year; the AI Application Software Index (LIST23492.US) rose 5.29% on the day, while the iShares Expanded Tech-Software ETF (IGV) gained 3.02%. Six AI hardware-related stocks were sold off:
Corning (GLW.US): Down 13%
SanDisk (SNDK.US): Down 10.62%
Micron Technology (MU.US): Down 10.57%
Intel (INTC.US): Down 9.03%
ASML (ASML.US): Down 4.64%
TSMC (TSM.US): Down 6.98%
Meta's Cloud Capital Background: $125B-$145B Capex and Three External Contracts
Meta's 2026 capital expenditure guidance of $125 billion to $145 billion was raised by $10 billion from the previous estimate during its Q1 earnings in late April 2026. The combined capital expenditures of the four major tech giants in 2026 are expected to reach approximately $725 billion, up 77% from the reported $410 billion in 2025. Unlike Microsoft (Azure), Google (GCP), and Amazon (AWS), Meta previously treated all infrastructure investments as pure cost expenditures with no corresponding cloud revenue to offset them.
To support massive AI computing demands, Meta signed three external contracts during the same period: a five-year, $60 billion strategic agreement with AMD to purchase custom Instinct GPUs; a $21 billion AI computing infrastructure contract with CoreWeave (announced expanded in April 2026, effective until 2032); and a $21 billion computing capacity contract with NEBIUS. The total value of the three contracts exceeds $100 billion.
A Sherwood News analysis in May pointed out that compared to other tech giants making similar large-scale investments, Meta lacks high-margin cloud business and enterprise revenue sources to offset capital expenditure impacts.
CoreWeave and NEBIUS Fall Over 10%: Three-Layer Market Logic of Largest Customer Becoming Potential Competitor
After the announcement, shares of CoreWeave (CRWV.US) and NEBIUS (NBIS.US) both fell more than 10%, with the market analysis revealing three layers of concurrent pressure.
Direct Competitive Threat: CoreWeave and NEBIUS's business model involves bulk purchasing of GPU clusters and reselling them at a markup, with high gross margins relying on tight AI computing supply. Meta enters the market with lower procurement costs from multi-billion-dollar strategic deals with NVIDIA and AMD, adding a large-scale participant on the supply side.
Identity Conflict: Meta is one of CoreWeave's largest customers, holding a $21 billion contract valid until 2032. Meta entering the same business creates overlapping roles of customer and competitor, leading market skepticism about contract renewal after 2032.
IPO Valuation Narrative Shaken: CoreWeave's IPO in March 2025 was based on the core valuation thesis of "explosive AI computing demand and extremely scarce supply." Meta's statement that it has the ability to sell excess computing capacity directly challenges the "supply scarcity" premise. CoreWeave reported Q1 2026 revenue of approximately $2.1 billion, with backlog supporting near-term revenue stability. UBS stated that the "overcapacity" narrative has raised concerns about actual AI demand. With Q2 and Q3 earnings seasons approaching, corporate capital expenditure guidance is key data for assessing whether the current valuation revaluation can persist.
Meta's Cloud Technology and Organizational Gaps: Key Points from Datafloq's June 2026 Analysis Report
Technology platform Datafloq noted in a June 2026 analysis report that Meta has been a pure B2C company since its founding, never selling products to enterprise customers, and lacks B2B sales experience. The report identified existing technology and organizational gaps: enterprise-grade multi-tenant isolation architecture, SOC 2/HIPAA/ISO 27001 compliance certifications, fine-grained billing and SLA guarantee systems, global multi-region deployment with network access points, and enterprise sales and customer success organizations.
The Datafloq report explicitly stated that "attempting to build a full-stack cloud platform would be a strategic mistake" and recommended focusing on specific areas, including raw compute leasing (hourly billing, API-orchestrated GPU clusters), Llama model hosting inference, and enterprise model fine-tuning services. According to Datafloq's analysis, Meta's recent cloud strategy resembles a "wholesale" computing sales model (similar to CoreWeave's model, targeting large customers with long-term contracts) rather than building a fully functional self-service cloud platform (like AWS). On the same day, Meta also established an "Enterprise Solutions" department, deploying engineers and product managers to large enterprise customers to assist with AI tool deployment.
Frequently Asked Questions
Why did Meta's stock rise over 6% after announcing its entry into cloud computing?
According to market analysis, the cloud business option transforms Meta's capital expenditure logic from "pure cost, one-way bet" to "a hedging option that can generate external revenue." At the May 2026 shareholder meeting, Zuckerberg publicly stated that cloud services were "absolutely under consideration" and that external companies inquire weekly about the possibility of purchasing Meta's computing power. The announcement came just five weeks after the shareholder meeting.
Why did Meta's entry into cloud computing cause AI hardware stocks to plummet?
Meta, one of the world's most aggressive AI computing investors, stated it may have the ability to sell excess computing capacity, directly shaking the core market assumption of "long-term scarcity in AI computing supply." This puts valuation pressure on semiconductor and memory chip makers that rely on tight supply-demand dynamics for high margins. Micron Technology and SanDisk each fell about 9%, and Corning fell 13%.
What are the specific market risks faced by CoreWeave and NEBIUS?
Meta, one of CoreWeave's largest customers with a $21 billion contract, entering the same business creates overlapping roles of customer and competitor. Both companies' IPO valuations are built on the "AI computing supply scarcity" narrative, which Meta's statement directly challenges. CoreWeave's Q1 2026 revenue was approximately $2.1 billion, with backlog supporting near-term revenue stability. Specific impacts on long-term contracts should be based on each company's official announcements.
