Google to partner with Marvell to develop new AI chips: Report

In a strategic move to enhance its artificial intelligence capabilities, Google is reportedly in negotiations with Marvell Technology to create two innovative chips. As detailed in a recent report by The Information, sources familiar with the discussions indicate that one of these chips may incorporate a memory processing unit designed to complement Google's existing tensor processing unit (TPU). The second chip is anticipated to be a new TPU optimized specifically for executing AI models. This initiative is part of Google's broader strategy to position its TPUs as a compelling alternative to Nvidia's dominant graphics processing units (GPUs). The increasing sales of TPUs have significantly contributed to the growth of Google's cloud revenue, as the company aims to reassure investors that its investments in AI are paying off. Furthermore, the report highlights that Google plans to finalize the design of the memory processing unit by next year, with subsequent steps leading to test production. This development is a testament to Google's commitment to enhancing its AI capabilities amid fierce competition in the tech sector. Recently, Google has been expanding its collaborations with other tech giants, including Intel and Broadcom, to strengthen its semiconductor ecosystem and improve chip design and manufacturing efficiency. As Google ramps up its efforts in AI accelerators, the potential emergence of a formidable competitor could challenge Nvidia’s long-standing market dominance. Nvidia, which has been working on its own AI inference chips leveraging technology from Groq, now faces intensified competition as Google enters the fray. This evolving landscape may reshape the AI chip market, posing new challenges to Nvidia's supremacy in high-performance computing. Google's first-quarter earnings report, set to be released on April 29, is anticipated to provide further insights into its AI investment strategy, cloud growth, and advertising performance, shedding light on how aggressively the company intends to compete in the semiconductor and AI sectors.