.png){kind=logo}
Google is preparing to introduce two new versions of its in-house artificial intelligence chips as it looks to strengthen its position in the growing AI computing market. According to a Bloomberg News report published on November 9, the company is developing customized versions of its Tensor Processing Units (TPUs) to meet different customer requirements.
The two variants, called TPU 8tx and TPU 8i Max, are not entirely new processors. Instead, they are reconfigured versions of Google's existing chip technology, designed to handle specific workloads, Google's Chief Technologist for AI Infrastructure, Amin Vahdat, told Bloomberg.
The TPU 8tx is designed for secure and confidential computing tasks. These workloads are particularly relevant for government agencies and businesses operating in regulated industries, where data protection and privacy are major concerns.
The second variant, TPU 8i Max, targets AI applications that require longer context windows. Such capabilities are important for AI systems that need to process and retain larger amounts of information while handling complex requests.
Also Read: Google’s Project Suncatcher Tests AI Chips in Space
Google's in-house TPUs have become an important part of its strategy to compete with NVIDIA, whose graphics processing units (GPUs) dominate the global AI computing market.
Initially developed for Google's internal requirements, these processors have gradually become an alternative for businesses looking to run AI workloads without relying entirely on NVIDIA hardware.
The latest development could further strengthen Google's position among companies building specialized AI infrastructure. However, Alphabet did not immediately respond to Reuters' request for comment, and Reuters said it could not independently verify Bloomberg's report.
Following the news, Alphabet shares gained more than 2% in premarket trading on Monday. However, the stock had declined around 9% since the beginning of 2026 through the previous trading session.