Snowflake announced on May 27, 2026 a multi-year strategic collaboration agreement with Amazon Web Services that includes the company's largest infrastructure commitment to date: $6 billion in Graviton compute and AI spend on AWS over five years. The expanded partnership is aimed at helping joint customers move enterprise AI from experimentation into production by bringing models closer to governed data.
$6 Billion for Graviton and AI Workloads
The $6 billion infrastructure pledge is structured as a multi-year commitment focused on AWS's Graviton general-purpose CPUs and GPU-accelerated EC2 instances for AI training and inference. Snowflake said the agreement reflects accelerating demand for data and AI workloads from its customer base, which has grown to more than 13,900 enterprises globally. Investors reacted strongly: the company's shares jumped roughly 36% in extended trading after the announcement and accompanying earnings beat.
Cortex AI Brings Models to Governed Data
At the technical core of the deal is Snowflake Cortex AI, which allows enterprises to run text-to-SQL, summarization, sentiment analysis, entity extraction and agentic workflows directly inside their Snowflake environment. "AI is only as powerful as the data behind it," Snowflake said, emphasizing that the architecture avoids moving sensitive information between systems.
Customers cited in the announcement include Fetch, which has deployed a Cortex AI semantic agent that lets sales teams query campaign data in natural language, and analytics platform Hex, whose co-founder said a secure and governed foundation is "what makes enterprise AI adoption real."
Marketplace, Regional Expansion and Industry Push
Snowflake on AWS Marketplace has crossed $7 billion in lifetime sales and topped $2 billion in 2025 alone, more than doubling year-over-year transaction growth. The new agreement extends joint go-to-market through AWS Marketplace with simplified contracting and faster procurement, and supports global expansion across 10 new AWS regions including New Zealand (Auckland), South Africa (Cape Town), Thailand (Bangkok) and the AWS European Sovereign Cloud.
The deal lands in a week dense with enterprise AI partnerships across the broader robotics and industrial AI ecosystem. FANUC and Google are bringing Gemini AI to 1.1 million industrial robots, while Catalyst Brands is tapping Figure AI humanoids for warehouses and Torc Robotics joined Mila to advance physical AI for trucks. In space, Schaeffler and Spire are partnering on a sovereign European satellite industry, underscoring how data and compute deals like Snowflake-AWS sit at the heart of the wider agentic AI buildout.
Why It Matters
The collaboration positions both companies to ride the next leg of enterprise AI: agents that act on governed data rather than chat with it. For Snowflake, deeper alignment with AWS shores up infrastructure economics ahead of Snowflake Summit 26 on June 1–4. For AWS, locking in the largest AI Data Cloud customer's compute spend is a meaningful counter to rival hyperscalers as the market for agentic workloads compounds.
Reporting based on coverage from Snowflake's press release, CNBC and TechCrunch.