Sign in Subscribe

Google Cloud Next 25: The New Way to Cloud - EXC Summary

Last updated on 
Google Cloud Next 25: The New Way to Cloud - EXC Summary
Google Cloud Next 25 event - EXC Summary

The Google Cloud Next 25 showcased Google's significant advancements and strategic direction, firmly positioning AI as the central pillar of its cloud offering.

It highlighted substantial investments in infrastructure, particularly AI compute with the announcement of the next-generation TPU "Ironwood" and advancements in quantum computing.

A major focus was placed on the introduction and expansion of Gemini AI models across Google Cloud and Workspace, demonstrating their increasing capabilities and integration into various products.

Furthermore, the eunveiled "Agents" as a transformative concept, showcasing how intelligent, autonomous systems are being built and deployed by customers to drive efficiency, innovation, and better user experiences. The emphasis was on providing an open, multi-cloud, and enterprise-ready AI platform that addresses concerns around security, privacy, and compliance.

Key Themes and Important Ideas/Facts:

 1. AI is the Core of Google Cloud's Future:

Google Cloud Next 25 event - AI is the Core of GCP's Future
  • Ubiquitous AI: Google is building the "most helpful AI" to transform ideas into enterprise solutions across various industries. Examples include agriculture, package inspection, power grid protection, scientific discovery, and customer service.
  • Significant AI Adoption: Over 4 million developers are using Gemini, with a 20x increase in Vertex AI usage. Google Workspace sees over 2 billion monthly AI assists.
  • Full Stack AI Investment: Google is investing in the entire AI stack, from infrastructure to research, models, products, and platforms.
  • $75 Billion Capex Investment: In 2025, Google plans to invest approximately $75 billion in capex, primarily directed towards servers and data centers to power AI compute and the cloud business.

2. Advancements in Infrastructure Powering AI:

Google's advancements in AI Infrastructure
  • Cloud Wide Area Network (WAN): Google's global private network is now available to enterprises, offering over 40% faster performance and up to 40% lower total cost of ownership. Sundar Pichai stated, "Today, I'm pleased to announce that we are making Google's global private network available to enterprises around the world. We call it Cloud Wide Area Network, or WAN."
  • Next-Generation TPUs (Ironwood): The seventh generation TPU, Ironwood, is launching later this year, achieving "3600 times better performance" than the first publicly available TPU. Pichai emphasized, "It's the most powerful chip we have ever built and will enable the next frontier of AI models."
  • Increased Energy Efficiency: Google has become 29x more energy efficient in the same period as the TPU advancements.
  • Quantum Computing Progress (Willow): Google's newest quantum chip, Willow, has cracked a key challenge in quantum error correction, paving the way for large-scale quantum computers.
  • AI Hypercomputer: A supercomputing system designed to simplify AI deployment, improve performance, and optimize costs, supporting various hardware platforms and a unified software stack.
  • Enhanced GPU Portfolio: Availability of A4X and A4 VMs powered by NVIDIA's Blackwell GPUs (GB200 and B200), and plans to offer NVIDIA's next-generation Vera Rubin GPUs.
  • Pathways for Cloud Customers: Google's distributed ML runtime powering Gemini is now available for cloud customers, enabling state-of-the-art, multi-host inferencing for dynamic scaling.

3. Gemini AI Models: Expanding Capabilities and Integration:

Gemini models and Google's Workspace Integration
  • Gemini 2.0: Introduced with advances in multimodality (native image and audio output) and native tool use.
  • Gemini 2.5 Pro: A "thinking model" with the ability to reason through its thoughts before responding. Sundar Pichai demonstrated its capabilities in solving complex reasoning challenges and creating physics simulations, noting, "This is a significant leap and shows the ability to produce robust, interactive code." It is available now in AI Studio, Vertex AI, and the Gemini app.
  • Gemini 2.5 Flash: A low-latency and cost-efficient model with "thinking built in," allowing control over the amount of reasoning for performance and budget balance.
  • Ubiquitous Integration: All 15 of Google's products with over half a billion users, including seven with over 2 billion users, are now powered by Gemini models.
  • Workspace Integration: Gemini is integrated into Gmail, Docs, Drive, and Meet, enhancing tasks like analysis in Sheets, audio overviews in Docs, and workflow automation.
  • Generative Media Models: Google offers a comprehensive suite of generative media models across all modalities available on Vertex AI. Thomas Kurian stated, "With Gemini, Imagen, Chirp, Lyria and Veo, Google is the only company that offers generative media models across all modalities..."
    • Models included:
      • Imagen 3: text to image
      • Chirp 3: custom voices
      • Lyria: text to music
      • Veo 2: video generation

4. The Rise of AI Agents:

  • Intelligent and Autonomous Systems: Agents are defined as intelligent systems with reasoning, planning, memory, and the ability to use tools to perform tasks under supervision.
  • Vertex AI Agent Builder and Agent Development Kit: Tools and frameworks for building sophisticated multi-agent systems, including an open-source Agent Development Kit, with support for the Model Context Protocol (MCP) and Agent2Agent protocols to enhance interoperability.
  • Google Agentspace: A platform that allows employees to find and synthesize information, converse with AI agents, and have them take action across enterprise applications. Gabe Weiss demonstrated Agentspace, showcasing its ability to automate tasks like portfolio analysis and meeting scheduling.
  • Agent Gallery: Within Agentspace, users can access a curated selection of purpose-built agents, including those powered by Google, the organization itself, and third-party models.
  • Customer Agent Examples - Simple Home and Garden Demo: Showcased a virtual customer service agent using voice and video to assist with product recommendations and landscaping services, including escalation to a human supervisor when needed.
  • Reddit: Using AI-powered search grounded in real conversations to enhance knowledge accessibility.
  • Retail: Implementing AI-powered search for product discovery and personalized recommendations.
  • Agents Categories
    • Customer (Customer Engagement Suite): Delight customers with personalized, human-like engagement from an end-to-end AI application that knows you, anticipates your needs, and delivers exceptional customer experience across every touchpoint
    • Creative Agents: Augmenting creative teams for content production at scale, including enriching experiences like the Wizard of Oz at the Las Vegas Sphere using Veo 2. Jim Dolan (CEO, Sphere) noted, "We ultimately came to the conclusion that Google was the only company that was actually capable of doing this."
    • Data Agents: Enabling data teams to manage and business teams to activate data, with specialized agents for data engineering and data science. Yasmeed Ahmad demonstrated how BigQuery, Colab, and Vertex AI powered by Gemini can simplify complex data analysis and forecasting.
    • Coding Agents: Like Gemini Code Assist, helping developers with code generation, debugging, and understanding code.
    • Security Agents: Integrated within Google Unified Security (GUS) for proactive threat detection, triage, and automated response. Payal Chakravarty and Nav Jagpal demonstrated GUS's ability to detect data leaks, malicious traffic, and recommend security enhancements.

5. An Open, Multi-Cloud, and Enterprise-Ready Platform:

  • Multi-Cloud Capabilities: Google Cloud offers an open multi-cloud platform that allows integration of AI agents with existing IT landscapes, including databases and applications, and interoperability with models and agents from other providers.
  • Enterprise Readiness: The platform is built for interoperability, addressing evolving concerns around sovereignty, security, privacy, and regulatory requirements. Thomas Kurian emphasized that customers can "adopt AI deeply while addressing the evolving concerns around sovereignty, security, privacy and regulatory requirements."
  • Comprehensive Grounding: Offering the most comprehensive approach to grounding model responses with Google Search, enterprise data, Google Maps, and third-party sources.
  • Strong Partner Ecosystem: Collaboration with ISVs, service partners (Accenture, Deloitte, etc.), and technology partners like NVIDIA to deliver integrated solutions and expertise. Jensen Huang (CEO, NVIDIA) stated, "Between NVIDIA and Google Cloud, this super partnership includes capabilities that covers literally every single layer and every single aspect of computing."
  • Sovereign Cloud Offerings: Enabling sovereign clouds with partners to meet international regulations, offering Google Cloud Sovereign AI services across public, sovereign, and distributed clouds, as well as with Google Workspace.

Quotes of Significance:

  • Thomas Kurian: "Today, that future [of AI for organizations] is being built by all of us."
  • Sundar Pichai: "With Google Cloud, we see AI as the most important way we can help advance your mission. The opportunity with AI is as big as it gets."
  • Amin Vahdat: "Demand for AI compute for training and inference is growing at an unprecedented rate. For over eight years, it has increased by over ten times, year over year, a factor of 100,000,000, just in eight years!"
  • Chris Kempczinski (McDonald's): "We're transforming our restaurant experience with the help of Google Cloud."
  • Jensen Huang (NVIDIA): "No company is better at every single layer of computing than Google and Google Cloud."
  • Gabe Weiss (Demonstrating Agentspace): "Agentspace is a game changer, and we can't wait to see how you put it to work."
  • Lisa O'Malley: "We are putting the power of Google Search into customer interactions."
  • Ynon Kreiz (Mattel): "We see Google Cloud as a true partner in bringing the magic of play to life for every Mattel fan."
  • Yasmeed Ahmad (Demonstrating Data Science Agents): "Gemini and Vertex AI have made BigQuery a complete data science platform, unlocking new insights faster than ever with natural language and code."
  • Payal Chakravarty (Demonstrating Google Unified Security): "GUS is an integrated, open platform that can protect any environment, any data from end point, viral, networks, identity, really any cloud, any model."

Conclusion:

The Google Cloud Next 25 clearly established Google's commitment to AI as the driving force behind its cloud strategy.

Significant investments in infrastructure, the continuous advancement and integration of Gemini models, and the introduction of powerful AI Agents across various domains signal a new era of cloud computing.

Google's emphasis on an open, multi-cloud, and enterprise-ready platform, coupled with a strong partner ecosystem, positions it as a key player in empowering organizations to innovate and transform with AI while addressing critical concerns around security and compliance. The numerous customer stories and product demonstrations provided compelling evidence of the real-world impact and potential of Google Cloud's AI-powered solutions.