Google Cloud Next 25 Unveils Powerful New TPU Chip and AI Agent Tools
Google announces a massive 20x surge in Vertex AI usage as over four million developers now build with Gemini. The company also unveils its most powerful chip yet, Ironwood, alongside new tools for creating AI agents.
Google showcases massive growth in artificial intelligence adoption at Google Cloud Next 25, revealing that over four million developers currently build with the Gemini model family. This rapid expansion drives a 20x increase in Vertex AI usage over the past year, while Google Workspace delivers over two billion AI assists monthly to business users. The company emphasizes its commitment to making AI more powerful, accessible, and affordable for global organizations.
A major highlight of the event is the introduction of Ironwood, Google's seventh-generation TPU designed specifically for AI inference. This new chip delivers five times the peak compute capacity and six times the high-bandwidth memory capacity compared to the previous generation. Additionally, Google integrates its Lyria model into Vertex AI, making Google Cloud the only platform to offer generative media models for video, image, speech, and music.
The tech giant also introduces robust new tools for building and managing AI agents, including the open-source Agent Development Kit (ADK) and the Agent2Agent protocol. Updates to Google Agentspace simplify the discovery and creation of AI agents, while the new AI Agent Marketplace allows customers to easily browse and purchase partner-built solutions. Google shares over 500 real-world examples of organizations achieving tangible business impact through these AI innovations.