Google Launches TPU 8tx and 8i Max, Ramping Up Its Challenge to NVIDIA

on Monday, November 9, Google launched two new variants of its Tensor Processing Units (TPUs), stepping up competition with Nvidia in the AI chip market. The TPU 8tx, aimed at government agencies and regulated industries, focuses on secure and confidential computing. This system operates alongside AMD chips as part of a partnership between the two companies; the TPU 8i Max is designed for AI applications with long context windows, combining two TPU chips into a single package to enhance compute performance and memory capacity.

Amin Vahdat, head of Google's chip business, stated that both new products are slated for release next year. Historically, Google has typically debuted one new AI processor every two years, but this year it introduced two simultaneously. Combined with this latest release, Google's TPU portfolio now spans four distinct models, each tailored to specific requirements. He explained that this strategy adheres to Google's consistent philosophy that the future belongs to specialization, and the competitive advantages derived from such focused development remain strikingly clear.

Industry analysts project that by 2028, sales related to Google's TPU ecosystem encompassing chips, software, and hardware will surpass $135 billion, significantly outpacing the roughly $12 billion estimated for this year. Upon news of the launch, Google's stock climbed more than 2% in early premarket trading, while Nvidia's premarket value dipped by nearly 6%.