Alphabet shares rose more than 2% in Monday premarket trading after Google introduced the TPU 8tx and TPU 8i Max, accelerating its challenge to Nvidia in specialized AI hardware. The TPU 8tx is designed for secure AI inference and smaller-scale model training by governments and regulated businesses, while the TPU 8i Max combines two TPU 8i processors to double computing power, onboard memory and high-bandwidth memory throughput within the same rack footprint. AMD supplies the CPU paired with the TPU 8tx, Broadcom has a longstanding chip-development relationship with Google, and Marvell Technology entered a custom-silicon partnership in August. Google has unveiled four TPUs in the past year, compared with its historical two-year development cycle. Oracle plans to offer Google TPUs through its cloud platform, and Anthropic, Apple and Meta are among companies using Google’s AI computing technology. Bloomberg Intelligence estimates TPU sales could exceed $135 billion in 2028, compared with approximately $12 billion this year, making customer adoption, manufacturing capacity and deployment timelines key metrics. The expansion creates another competitive challenge for Nvidia, while Google’s ability to scale production and deliver competitive performance will determine whether its faster product cycle produces meaningful market-share gains.