Meta Platforms Inc.Meta plans to manufacture Iris in September, aiming for cost control and greater control over computing stack, but execution risk remains.
Meta Platforms plans to begin manufacturing its in-house AI chip, code-named Iris, in September, working with Broadcom on design and Taiwan Semiconductor Manufacturing Company on production. The company also plans to deploy seven gigawatts of computing capacity in 2026 and reach 14 gigawatts in 2027. Iris is meant to complement, rather than immediately replace, the large quantities of GPUs Meta buys from Nvidia and AMD, aiming for cost control and greater control over a computing stack that supports content recommendations, advertising tools, and AI features across Facebook and Instagram. Custom silicon could strengthen the economics of Meta's core moat—its family of social platforms, network effects, and ad-targeting infrastructure—if it lowers inference costs at scale, though infrastructure spending remains massive and custom chips require sustained execution.
Meta Platforms Inc.Meta plans to manufacture Iris in September, aiming for cost control and greater control over computing stack, but execution risk remains.
Broadcom IncMeta is working with Broadcom on the design of Iris, generating design service revenue.
NVIDIA CorporationMeta's custom chip Iris may reduce reliance on Nvidia GPUs, though it complements rather than replaces them for now.
Advanced Micro Devices IncMeta's in-house AI chip Iris may reduce reliance on AMD GPUs, potentially lowering future demand.
Taiwan Semiconductor Manufacturing Co. Ltd.TSMC will manufacture Meta's Iris chip, adding to its production orders.