Nvidia rises about 2% after Meta readies in-house AI chip
Nvidia shares rose about 2% Friday after reports Meta will start producing an in-house AI chip called Iris in September; Morgan Stanley kept an Overweight rating and a $288 target.
Nvidia shares climbed about 2% on Friday, trading near $207 after reports that Meta plans to begin producing an in-house artificial intelligence processor called Iris in September. Morgan Stanley reiterated an Overweight rating and a $288 price target, and TD Cowen reaffirmed a Buy rating with a $275 target.
The Iris processor is part of Meta’s Meta Training and Inference Accelerators program and is intended to support AI systems that run Facebook and Instagram. Company testing of the chip lasted six weeks and found no major issues.
Meta is working with Broadcom on Iris’s design and plans to have Taiwan Semiconductor Manufacturing Co. produce the chips. Company documents indicate the processor is meant to augment the large volumes of graphics processing units Meta continues to buy from Nvidia and Advanced Micro Devices rather than replace them.
Meta has developed several MTIA generations over more than five years. To date, the custom chips have mainly been used for inference — running trained models to generate outputs — while earlier MTIA designs were aimed at expanding into AI training workloads over time.
Analysts maintained positive views on Nvidia following recent meetings and a non-deal roadshow with company executives. Morgan Stanley described Nvidia as presenting an accelerating and more diversified growth profile. TD Cowen reported that Nvidia executives outlined sustained demand for AI computing, constrained availability of compute capacity, rising rental prices for older GPUs, expanding enterprise AI adoption, and cloud agreements at premium pricing.
On the trading session, Nvidia shares recovered from weaker premarket activity and were up about 2.3% at roughly $207. The development follows other major technology firms that have invested in custom chips to reduce computing costs and alter parts of their procurement strategies.








