# Google Unveils TPU 8tx and TPU 8i Max, Two New AI Chips for Confidential Training and Bigger Models

*2026-11-14 11:00 UTC · AI · 1 min read*

> Google added two eighth-generation TPUs, one built with AMD for confidential training and inference, the other doubling inference memory and bandwidth.

Google on Monday unveiled two new variants of its in-house tensor processing units (TPUs), widening its chip portfolio to confidential AI workloads and memory-hungry models. Both belong to the eighth TPU generation introduced in April, as Google works to reduce its cloud business's dependence on chips made by others.

TPU 8tx, designed with AMD and based on the TPU 8t, targets confidential training and inference. Google says it builds security into the hardware layer, keeping sensitive data and model weights isolated even while in use. A custom inferencing network lets it run reasoning alongside training, so companies can distill and train small models on proprietary data without moving it between devices.

TPU 8i Max combines two of the inference-focused TPU 8i chips from April in one package, doubling memory capacity, processing power and bandwidth so models can handle more context per query. The launch comes as Microsoft, Amazon, OpenAI and Anthropic also develop their own AI chips to loosen Nvidia's grip, days after Jensen Huang's resignation as Nvidia CEO.

## Sources

- [techcrunch.com](https://techcrunch.com/2026/11/10/google-broadens-its-ai-hardware-portfolio-with-two-new-specialized-chips/)

---
TechTools News · https://techtools.cz/tools/news/article/2687-google-unveils-tpu-8tx-and-tpu-8i-max-two-new-ai-chips-for-confidential
Česká verze: https://techtools.cz/tools/news/clanek/2761-google-predstavil-tpu-8tx-a-tpu-8i-max-dva-nove-ai-cipy-pro-duverne-trenovani-a