Google Cloud Unveils AI Chips

At its annual Cloud Next conference, Google Cloud unveiled a significant expansion of its AI infrastructure with the launch of two new specialized chips.

The TPU 8t is optimized for high-speed model training, while the TPU 8i is dedicated to efficient inference workloads, marking a strategic split from previous all-in-one generations.

Google also introduced the Gemini Enterprise Agent Platform, a suite of tools intended to help businesses build and manage autonomous AI agents.

These developments signal Google's commitment to leading the 'agentic era' of computing while reducing reliance on third-party hardware providers.