Full Breakdown
Meta Moves Its Custom AI Chip “Iris” to Production and Plans to Double Data-Center Capacity
7/18/2026, 11:38:21 AM
Core Event: Iris Chip Production and Data-Center Expansion
An internal memo disclosed that Meta Platforms will begin production of its custom AI ASIC, Iris, in September 2026. The company also announced a plan to raise its data-center power capacity from the current level to 14 gigawatts by 2027, effectively doubling the infrastructure needed for AI workloads.
Background & Context: AI-Centric Capital Strategy
Meta’s CEO Mark Zuckerberg has positioned AI as the engine for future revenue growth, especially for the advertising business that serves over 3.5 billion daily active users. Earlier in the year, CFO Susan Li warned that data-center capacity projected 12–36 months ago is now insufficient, creating a bottleneck as demand for AI processing rises. To address this, Meta partnered with Broadcom to design Iris and will have the chip fabricated by Taiwan Semiconductor Manufacturing. The move follows a broader industry trend in which the “top four hyperscalers”—Meta, Microsoft, Amazon and Alphabet—are slated to spend $600 billion to $700 billion on capital expenditures in 2026.
Data & Statistics: Investment Scale and Financial Impact
- Planned capital outlay for 2026: up to $145 billion.
- Targeted data-center capacity: 14 GW in 2027.
- Q1 2026 revenue growth: +33 % YoY, driven largely by AI-enhanced ad performance.
- Forward price-to-earnings multiple: 21, below the roughly 25 multiple of other “Magnificent Seven” peers.
- Gross margin: 81.94 %, the highest among the Magnificent Seven.
- Trailing cash flow from operations: $124 billion, providing liquidity for AI initiatives.
Official Statements & Responses
During the first-quarter earnings call, Zuckerberg said, “One of the primary goals of our Meta Compute initiative is to lead the industry in efficiency of building compute, and we expect that will be a strategic advantage over time.” The company framed the Iris rollout and capacity expansion as essential steps to lower AI-compute costs, improve recommendation systems, and sustain ad-revenue growth.
Verbatim Quotes
- “One of the primary goals of our Meta Compute initiative is to lead the industry in efficiency of building compute, and we expect that will be a strategic advantage over time,” — Mark Zuckerberg, CEO, Meta Platforms
- “Meta has the highest gross margin of any Magnificent Seven company.” — The Motley Fool analysis
- “Its $124 billion in trailing cash flow from operations is a strategic advantage, helping fund its AI initiatives.” — The Motley Fool analysis
- “The Motley Fool's research shows that the top four hyperscalers -- Meta, Microsoft, Amazon, and Alphabet -- plan to spend between $600 billion and $700 billion on capex in 2026.” — The Motley Fool analysis
These developments position Meta to compete more aggressively in the AI infrastructure space and could prompt a re-rating of its stock if the expanded compute capacity translates into higher profitability.
