Snowflake Inc.’s shares jumped in late trading today after it announced a $6 billion spending commitment on Amazon Web Services Inc.’s cloud infrastructure, including a deal to use the company’s custom artificial intelligence chips.

The cloud data platform giant also reported stellar first-quarter earnings results, powering past Wall Street’s targets in a sign that the AI boom is providing it with a real tailwind, contrary to many investor’s fears.

The company’s purchase of AWS’ cloud technology and services will span the next five years. It includes a commitment by Snowflake to use more of Amazon’s general-purpose Graviton chips, as well as its custom AI accelerators.

For AWS, the deal underlines the increasing momentum it’s gaining in the AI industry, as an increasing number of businesses turn to its trusted cloud platform to run more sophisticated “agentic” AI workloads. It follows an announcement by Claude chatbot creator Anthropic PBC last month, which revealed it will spend $100 billion on AWS infrastructure over the next decade. AWS has also struck a multibillion-dollar deal with the rival AI firm OpenAI Group PBC.

The deals with Anthropic and OpenAI include equity investments, but Snowflake is a more mature and publicly traded company, so that’s not the case in today’s deal. Snowflake, which went public back in 2020, boasts a market capitalization of more than $60 billion, and has long relied on Amazon’s cloud services.

Snowflake’s latest financial results were compelling. The cloud data giant reported earnings before certain costs such as stock compensation of 39 cents per share, cruising past the Street’s target of 32 cents. Meanwhile, its revenue soared by 33% from a year ago, to $1.39 billion, surpassing the $1.32 billion consensus estimate by a solid margin.

Snowflake’s guidance was strong too. The company called for an adjusted operating margin of 12.5% and between $1.415 billion and $1.42 billion in second-quarter product revenue. Wall Street is looking for a margin of just 11.9% on product revenue of $1.37 billion.

The company also raised its full-year revenue guidance. It’s now looking for total product revenue of approximately $5.84 billion, up from an earlier call for $5.66 billion. That puts it well ahead of the analyst’s target of $5.67 billion. According to Snowflake Chief Executive Sridhar Ramaswamy (pictured), the company decided to raise its guidance as a result of the “strong momentum” it’s seeing both in its core business and in AI.

“AI continues to be a powerful tailwind for Snowflake, and Q1 marks a clear inflection point in that journey,” he said. “With Cortex Code and Snowflake Intelligence, we are extending from the trusted foundation for enterprise data and context to become the control plane for the Agentic Enterprise. We are seeing strong momentum from both AI-driven acceleration of our core platform and growing adoption of our first-party AI products, positioning Snowflake to lead in this new era.”

The report comes in the wake of a rough few months for software companies, during which investors have become increasingly concerned that AI startups could be about to cause some major disruption to the old way of doing things in the enterprise. Many are convinced that AI agents, which can complete tasks based on a simple, plain language prompt, will completely reshape the way business gets done.

One of the main fears is that enterprises will simply use coding agents to build customized versions of the software they currently pay hefty subscription fees for. In an earnings call earlier this month, executives of Palantir Technologies Inc. claimed to have done this, replacing their customer relationship management platform with an agentic-built system.

Agents may also pose problems for software firm’s business models. As they become more numerous on corporate networks, that’s going to clash with the per-user fee structure that many software companies operate, forcing them to switch to consumption-based pricing instead.

However, Snowflake already does this, charging its customers based on the amount of storage and compute resources they use, though its gross margins are generally quite a bit lower than software firms that run on user-based pricing models. With Snowflake’s cloud data warehouse, companies pool all of their data from multiple different sources into one centralized repository, where it can serve as a clean and governed source of truth for enterprise data analytics and forecasting.

Snowflake has long argued that its software won’t be replaced by AI, but instead will actually become invaluable for it, acting as a hub that enables autonomous agents to work more reliably. After all, agents are no different to human workers in that they still need a way to access cleaned and governed enterprise data. Where they do differ is that they can work much faster, which could increase consumption of Snowflake’s cloud and boost its bottom line.

The company also develops its own AI agents, and they are the reason behind its acquisition of a startup called Natoma, which was also announced today. Natoma is an enterprise-grade Model Context Protocol platform for AI agents, and its technology will enable Snowflake to build a natively integrated governance and identity layer for AI agents that require MCP tool access. Using this layer, enterprises will be able to securely connect AI agents to third-party databases, APIs and software platforms, the company explained.

Investors were extremely encouraged by what they saw, and Snowflake’s stock gained more than 36% in late trading, helping to erase much of the losses it has suffered this year. However, Snowflake still has a way to go, for its stock is still down 20% in the year to date, while the broader S&P 500 Index has gained 10% over the same timeframe.

Photo: Robert Hof/SiliconANGLE