Meta is on track to start making the latest versions of its AI‑specific chip in September, aiming to lower GPU costs amid a component shortage.

Meta announced that its next‑generation AI‑specific chips will enter production this September, a move aimed at easing GPU price pressure as the industry grapples with a component shortage.

Production timeline and capacity

The company plans to begin silicon fabrication at its partner foundry in early September, with initial volume shipments slated for the fourth quarter. Meta expects the new chips to supplement its existing AI hardware fleet, providing additional compute capacity for internal services and external developers.

Design focus on cost efficiency

Unlike prior generations that emphasized raw performance, the upcoming chips prioritize lower manufacturing costs and energy consumption. Meta’s engineering team has integrated a streamlined architecture that reduces the number of high‑end GPU cores while maintaining sufficient throughput for common machine‑learning workloads.

Impact on the AI hardware market

If the chips meet Meta’s cost targets, they could pressure other AI hardware vendors to revisit pricing strategies, especially as demand for affordable compute continues to outpace supply.

  • Reduced reliance on external GPU providers
  • Potential price drops for cloud‑based AI services
  • Increased accessibility for smaller AI startups

Meta has not disclosed detailed specifications, but industry analysts anticipate that the chips will support mixed‑precision training and inference, aligning with the company’s broader push toward on‑device AI.

"Our goal is to democratize AI access by delivering high‑performance, cost‑effective silicon," a Meta spokesperson said in a statement.

Read the full coverage at TechCrunch.