Shares of Meta Platforms (NASDAQ:META) surged 5.2% during afternoon trading today as investors recalibrated the company’s massive capital expenditure, viewing its aggressive AI infrastructure investments as a catalyst for efficiency, new revenue streams, and accelerated growth rather than a mere drain on resources.
Custom Silicon: Cutting Dependence on Nvidia
A major driver of the market’s optimism is Meta’s move to lower operational costs through custom hardware. The company confirmed it will begin manufacturing its proprietary “Iris” AI chip this September in partnership with Broadcom and TSMC. By taking control of more of the silicon stack, Meta aims to reduce its reliance on expensive GPUs from Nvidia and AMD, potentially lowering the long-term costs of both training and inference.
Monetizing the Infrastructure Gap
Meta is actively transforming its data center footprint into a commercial asset. With a $9.1 billion (CAD 13 billion) facility underway in Canada and a target of 14GW of compute capacity by 2027, the company is positioning itself to enter the cloud market. By renting out excess capacity as an “AI Cloud,” Meta is turning idle infrastructure into a viable revenue stream. Furthermore, the company’s ability to build capacity at costs significantly below the industry average of $45 billion-per-GW provides it with superior unit economics compared to many of its competitors.
Aggressive Growth and Ecosystem Expansion
Meta is also moving to dominate the developer space with its new coding model, Muse Spark 1.1. By pricing the model at approximately one-quarter of the cost of competing offerings from OpenAI and Anthropic, Meta is making a strategic play to pull developers into its ecosystem. This strategy is designed to widen the funnel for future tool integration, advertising demand, and cloud usage.
Execution Risks and Market Outlook
Despite the bullish sentiment, the company’s success hinges on execution. Custom silicon and multi-gigawatt builds only deliver value if developer adoption, margins, and utilization rates align with the massive upfront spend. Until these metrics are reflected in financial results, the market remains wary of the potential for overbuilding.
Today’s 5.2% jump follows a similar trend from nine days ago, when the stock gained 10.1% on reports that Meta plans to launch a cloud business to sell surplus AI computing capacity. This internal initiative, referred to as “Meta Compute,” aims to offer both Model-as-a-Service hosting and raw compute, effectively positioning Meta as the fourth U.S. hyperscaler alongside AWS, Azure, and Google Cloud.
This pivot addresses the primary concern that has weighed on Meta’s stock throughout the year: the ballooning capital expenditure. With 2026 AI spending projected between $125 billion and $145 billion, investors previously feared for the company’s free cash flow. By framing this spending as the foundation for a new cloud business, Meta is attempting to turn a feared cost center into a core revenue pillar.
Meta is currently up 2.6% year-to-date. Trading at $667.43, the stock remains 15.5% below its 52-week high of $790, reached in August 2025. Long-term performance remains strong, with a $1,000 investment from five years ago currently valued at $1,890.