Meta’s stock rebounds as AI coding and custom chips ease spending fears

Meta just pulled off something rare in the current tech landscape: convincing Wall Street that spending more money is actually a good thing.

The company’s shares climbed roughly 8% after reports emerged that Meta is exploring a cloud business to sell excess AI computing power. Major tech firms collectively shed over $1 trillion in market capitalization in a single week as investors grew increasingly nervous about the sheer scale of AI infrastructure spending.

Custom silicon and the Iris chip

At the center of Meta’s rebound story is a chip called “Iris.” Production is scheduled to begin in September 2026, marking Meta’s first custom data center AI chip.

Iris falls under Meta’s broader Meta Training and Inference Accelerators (MTIA) project, an initiative designed to give the company more control over its AI computing stack. Broadcom handles chip design, TSMC handles manufacturing, and Meta has committed to a 1-gigawatt deployment of custom AI chips through this Broadcom partnership extending to 2029.

For a company that just raised its 2026 capital expenditure guidance to between $125 billion and $145 billion, finding ways to make that spending more efficient isn’t just nice to have. That capex range is designed to expand Meta’s computing capacity to 14 gigawatts within the next year.

The cloud play that changed the narrative

Meta is building massive AI computing capacity for its own needs, but there will inevitably be periods when some of that capacity sits idle. Rather than letting expensive hardware gather dust, Meta would sell surplus compute to outside customers.

The roughly 8% stock jump on July 1 following news of this potential venture tells you how hungry the market was for any signal that Meta’s AI spending had a monetization path beyond its core advertising business.

Why this matters beyond Menlo Park

The trillion-dollar wipeout in tech market cap over a single week underscores just how fragile investor confidence has become around AI infrastructure spending. Meta’s ability to articulate a clearer path to monetization, through both custom chips and cloud services, is part of what separated it from the pack during this selloff.

Meta’s partnership with TSMC for manufacturing its Iris chips adds another major customer competing for advanced chip fabrication capacity.

The $125 billion to $145 billion capex range is intended to support Meta’s advertising business, its metaverse ambitions, and potentially its new cloud operations. The 1-gigawatt custom chip deployment commitment through 2029 with Broadcom gives investors a concrete timeline to measure progress against. Key metrics to track include execution timelines on the Iris chip production starting in September, any formal announcements around the cloud division’s structure and pricing, and whether Meta’s AI-driven ad revenue growth can keep pace with its infrastructure spending.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.