Meta to Deploy Homegrown Chips by 2027

Tech giant plans four new generations of in-house AI chips to power expanding AI workloads.

Published on Mar. 11, 2026

Meta, the parent company of Facebook, plans to deploy four new generations of its own custom-designed AI chips by the end of 2027 to help power the company's rapidly expanding artificial intelligence workloads. This move comes as demand for AI computing continues to surge, with companies like Oracle also seeing strong sales in this area.

Why it matters

Meta's shift to homegrown chips signals the company's growing reliance on AI and machine learning technologies to power its various products and services. Developing its own silicon gives Meta more control over the performance and capabilities of the hardware underlying its AI systems, which are becoming increasingly central to the company's operations.

The details

Meta will roll out four new generations of its in-house AI chips over the next five years to support the company's growing AI workloads. This includes powering features like content recommendations, language processing, and computer vision across Meta's family of apps and services. The move allows Meta to optimize chip design for its specific AI needs rather than relying on off-the-shelf processors.

  • Meta plans to deploy the four new chip generations by the end of 2027.

The players

Meta

The parent company of Facebook, Instagram, and other social media platforms, which is investing heavily in artificial intelligence technologies.

Oracle

A major technology company that has also seen strong sales growth in the AI computing market.

Uber

The ride-hailing company that is partnering with Zoox to offer autonomous vehicle services in Las Vegas.

Zoox

An autonomous vehicle company owned by Amazon that is providing self-driving cars for Uber's new robotaxi service in Las Vegas.

Got photos? Submit your photos here. ›

What they’re saying

“Demand for AI computing shows no sign of letting up.”

— Caroline Hyde, Bloomberg Reporter (Bloomberg)

The takeaway

Meta's move to develop its own AI chips underscores the growing importance of artificial intelligence across the tech industry, as companies race to build custom silicon to power increasingly sophisticated machine learning applications.