Google's latest move to bolster its artificial intelligence hardware efforts involves a partnership with Marvell Technology to develop two new chips, according to a recent report. This collaboration aims to enhance efficiency and directly challenge Nvidia Corp.'s dominance in the AI chip market.

The reported agreement involves the creation of a memory processing unit (MPU) intended to complement Google’s existing Tensor Processing Unit (TPU). Alongside the MPU, Google and Marvell are reportedly developing a new, specialized TPU designed specifically for optimizing the execution of AI models. This signifies a significant investment in Google’s TPU strategy, signaling its commitment to building a robust in-house AI hardware ecosystem.

The move highlights the escalating competition within the AI chip industry. Nvidia currently holds a substantial lead in the market, fueled by high demand for its GPUs used in training and deploying AI models. Google's increased focus on its own AI chips, particularly TPUs, represents a strategic effort to reduce reliance on third-party providers and gain greater control over its AI infrastructure. The development of a dedicated MPU alongside the TPU suggests a holistic approach to optimizing AI workloads, addressing both memory bandwidth and processing power bottlenecks.