You won't get into the SpaceX IPO (From Behind the Markets)

Google Cloud Next 2026 wrapped up its opening two days in Las Vegas with one of the most ambitious and wide-ranging sets of announcements in the event's history.

For investors watching Alphabet (NASDAQ: GOOGL), the timing is significant. The stock closed on Thursday at $338.89, just 2.9% from its all-time high. And with Q1 2026 earnings due April 29, the company just gave the market a detailed preview of exactly what its $175 billion to $185 billion 2026 capital expenditure (CapEx) commitment is actually building.

The answer is a full-stack AI enterprise platform, and the announcements from Cloud Next make the vision more concrete. Here are some of the event’s biggest announcements and takeaways so far.

The 8th-Generation TPU: 2 Chips for the Agentic Era

The headline infrastructure announcement was the unveiling of Google's eighth-generation TPUs, a dual-chip architecture designed specifically for what the company calls the agentic era. The TPU 8t is optimized for accelerated model training, while the TPU 8i targets cost-effective inference at near-zero latency. As AI deployments shift from training large models to running them continuously at enterprise scale, inference cost and latency become the primary competitive variables. Google's decision to build dedicated silicon for each workload reflects a level of infrastructure sophistication that is difficult to replicate.

Alongside the TPUs, Google announced the Virgo Network, a new megascale data center fabric designed to underpin the AI Hypercomputer, its term for clusters of thousands of interconnected chips operating as a single system. Managed Lustre storage, delivering 10 terabytes per second of throughput, rounds out the infrastructure stack. Together, these announcements signal that the $175 billion to $185 billion CapEx commitment is being deployed with precision, and not just scale.

Vertex AI Is Now the Gemini Enterprise Agent Platform

Perhaps the most strategically significant announcement was the rebranding of Vertex AI as the Gemini Enterprise Agent Platform. The rebrand signals a fundamental shift in how Google is positioning its cloud business. This is no longer a collection of machine learning tools. It is a unified, end-to-end control plane for building, deploying, securing, and orchestrating AI agents at enterprise scale.

Google Cloud CEO Thomas Kurian framed the competitive argument directly: while rivals are handing customers pieces, Google is offering the whole platform. It is a bold claim, but the logic has merit. Google is the only major hyperscaler that controls custom silicon, frontier AI models, a cloud platform, and an enterprise productivity suite with billions of users. The Gemini Enterprise Agent Platform is the layer that attempts to unify all four into a single operating system for the AI enterprise.

The platform includes a new Agent Designer for building schedule and trigger-based agents, long-running agents capable of executing complex business processes, and an inbox for managing agent activity, all integrated natively with Google Workspace.

Workspace Intelligence and the Enterprise Distribution Advantage

Workspace Intelligence reached general availability at Cloud Next, delivering, according to Google, a unified, real-time understanding across its productivity applications. The system goes beyond simple data retrieval, incorporating a dynamic understanding of semantic relationships across documents, projects, collaborators, and organizational context. Rapid Enterprise Migration, also announced, now enables organizations to migrate from Microsoft 365 to Google Workspace up to five times faster than before, a direct competitive thrust against Microsoft's (NASDAQ: MSFT) dominance in enterprise productivity.

The scale of Google's existing distribution through Workspace is often underappreciated in discussions of the AI cloud race. Three billion users across Workspace apps represent a deployment channel that neither Amazon's (NASDAQ: AMZN) AWS nor Microsoft's Azure can match through productivity software alone.

The Numbers Behind the Conference

CEO Sundar Pichai revealed that 75% of all new code written at Google is now AI-generated, up from approximately 25% just a year ago.

Google's first-party models are now processing 16 billion tokens per minute via direct customer APIs, up from 10 billion in the prior quarter. Google also committed $750 million to a partner fund to accelerate the adoption of agentic AI across its 120,000-member global partner ecosystem.

Analysts hold a consensus Moderate Buy rating on GOOGL, with a price target of $369.67, implying about 7% upside potential.

Q1 earnings on April 29, where Google Cloud growth is expected to exceed 50% year over year, is the next test of whether the infrastructure investment is translating into revenue at the pace the market is pricing in.

Read this article online ›