Meta Rises Over 8% on Report Company Is Selling Surplus AI Computing Power

TradingKey - On July 1, Eastern Time, Meta Platforms ( META) opened with a rapid surge in its share price, rising over 8% to mark its largest intraday gain in nearly two months. As of press time, Meta was up 7.74% at $606.88.

[Source: Futu]

In terms of news, media reports indicate that Meta is building a cloud business and plans to sell its surplus AI computing power and model access. This model benchmarks against AWS and Azure, meaning Meta could unlock a brand-new, high-margin revenue stream.

According to people familiar with the matter, Meta has acquired a massive amount of data centers and chips for this purpose, planning to sell access to AI models (including its self-developed Muse Spark) hosted on its existing infrastructure and charge developers. If the plan is implemented, it will mark the first time Meta systematically offers its internal infrastructure to external users, directly entering the enterprise cloud services market.

Currently, Meta has not released an official announcement on the matter. In a statement sent to the media, the company said: "We are always exploring ways to utilize our infrastructure resources more efficiently, but we have no specific information to share at this time."

Whether or not the cloud business ultimately materializes, Meta's investment in infrastructure is already a done deal. In April this year, the company raised its 2026 capital expenditure guidance to $125 billion to $145 billion, with the vast majority destined for AI data centers and computing chips. In February this year, AMD announced a five-year partnership agreement with Meta to deploy 6GW of computing capacity, with the first 1GW expected to be delivered in the second half of 2026.

In addition, Meta has locked in approximately 1.6GW of computing capacity from data center developer Crusoe Energy. In terms of reserve scale alone, Meta is on par with other cloud providers.

Against the backdrop of increasingly scarce AI computing resources, Meta's potential move into the cloud business has drawn significant market attention. Since the beginning of this year, leading AI companies like OpenAI and Anthropic have repeatedly complained publicly that GPU shortages are limiting the training and inference capabilities of large models. However, because Meta's internal workloads have not yet fully consumed its massive procurement reserves, its vast computing inventory has instead become leverage for external monetization. Previously, it was reported that Google ( GOOGL) failed to meet Meta's demand to access Gemini due to tight supply, as computing capacity bottlenecks reshape the industry's cooperative landscape.