Google’s New, End-to-End AI Computing Ecosystem for Science, Engineering, and Technical Computing

About the Session

The boundaries of scientific exploration, engineering design, and technical computing are being rewritten by the rapid evolution of AI capabilities. Researchers and engineers are no longer just using AI as passive chatbots; they are deploying full-scale, end-to-end agentic systems capable of generating novel ideas and hypotheses, orchestrating complex multiphysics simulations, generating production-ready code, and reasoning through massive scientific datasets. Google has recently announced important new technologies and upgrades at Google Cloud Next and Google I/O that will radically augment human capabilities and enable powerful automation capabilities.

Join the Google Cloud Advanced Computing Community for an all-encompassing overview of Google’s complete AI technologies explicitly tailored for technical and scientific domains–and how they complement each other. This session will serve as a comprehensive map of the newly updated ecosystem with a focus on scientific & technical computing, demonstrating how our models, developer tools, and enterprise agent platforms integrate seamlessly with your existing workloads to accelerate the time-to-insight, innovation, or decision.

The Complete AI Stack We Will Explore Includes Several New/Updated Technologies:

• The Next-Gen Foundation Models: Get an inside look at the brand-new Gemini 3.5 model family, precision-engineered for the agentic era with radical improvements in rapid execution loops, long-horizon tasks, complex reasoning, and technical coding cycles.

• Agentic AI from Ideation to Optimization: Learn how we tailor this core intelligence for advanced research using specialized initiatives like Gemini for Science (including Co-scientist and AlphaEvolve) 

• Gemini Spark to bring cognitive capabilities directly into your documentation and collaborative environments.

• Agent-First Developer Workspace: Discover Google Antigravity 2.0, the newly rebuilt environment designed entirely around agentic execution. Learn how to leverage the new Antigravity CLI, SDK, and specialized Science Skills to let autonomous sub-agents debug code, compile software asynchronously, and manage local technical workflows without blocking your terminal.

• Enterprise-Grade Agent Platforms: Step into the Gemini Enterprise Agent Studio and Agentic Platform to see how we provide a secure, low-code interface to build, test, and schedule multi-day autonomous workflows.

• The Agentic Data Cloud: See how the modern data ecosystem utilizes a semantic Knowledge Catalog to let agents safely reason across both structured simulation data and unstructured research repositories.

Whether you are a technical applications developer looking to drastically cut down debugging cycles, a data scientist managing petabyte-scale pipelines, or a lab director aiming to implement automated "AI Co-Scientist" workflows, this session will show you how to securely harness the full power of Google AI for advanced technical computing.

Speaker