Google is in talks with Marvell Technology to develop a memory processing unit that works alongside TPUs, and a new TPU for running AI models

Google is in talks with Marvell Technology to co-develop two new chips focused on making AI model execution more efficient. The details come from two people with direct knowledge of the discussions.

One chip would be a memory processing unit designed to work alongside Google's existing TPUs. The other would be an entirely new TPU built specifically for running AI models.

No deal has been finalized yet — these are still active discussions. But the move signals Google's continued push to build custom silicon for its AI workloads rather than relying solely on off-the-shelf hardware.

Marvell, already a major player in custom chip design for cloud giants, would gain another heavyweight partner if the talks materialize into a formal agreement.