MENLO PARK, California, May 27, 2026, 14:04 PDT
- Snowflake lifted its forecast for fiscal 2027 product revenue after first-quarter revenue beat Wall Street expectations.
- The company is putting $6 billion into AWS infrastructure over five years, with the money going to Graviton processors and AI capacity.
- Shares moved higher after hours. Investors seemed to shrug off concerns that AI agents might hit regular software spending.
Snowflake lifted its full-year product revenue outlook Wednesday and inked a $6 billion, five-year cloud infrastructure agreement with Amazon Web Services. That further links the data software firm to the AI investment surge. The stock surged 29% after hours, Reuters said.
Investors want signs that artificial intelligence is adding to revenue for software firms and not only raising costs. Snowflake, which provides data storage, analytics and sharing tools, has seen more of that data going toward building AI systems in big companies.
Snowflake reported first-quarter product revenue up 34% to $1.33 billion and total revenue up 33% to $1.39 billion for the period ended April 30. The company also raised its product revenue outlook for fiscal 2027 to $5.84 billion, up from $5.66 billion.
Snowflake’s new AWS deal is its biggest infrastructure commitment yet with Amazon’s cloud. The company said it plans to spend on Graviton compute — Amazon’s Arm chips — and AI gear over five years, and will keep growing joint efforts on generative AI and agentic AI. These AI systems can act and manage tasks, not just return answers.
Snowflake CEO Sridhar Ramaswamy said the company is entering “the era of the agentic enterprise,” describing a shift where AI systems help businesses use trusted data to drive outcomes. AWS chief Matt Garman pointed to Snowflake’s deeper adoption of Graviton chips, saying it should bring the cost, performance and flexibility customers need to run data warehousing and AI workloads at scale. Snowflake
Snowflake reported over 13,600 accounts tapped its AI features in the quarter. More than 7,100 accounts used Cortex Code. The company also said Snowflake Intelligence accounts more than doubled from the previous quarter.
Snowflake CFO Brian Robins said the company has 779 customers spending at least $1 million over the past year, with 46 of those reaching that level last quarter. Robins said growth in data and AI units led to the increased full-year outlook.
Snowflake projected second-quarter product revenue of $1.415 billion to $1.420 billion, topping the $1.37 billion analyst consensus from LSEG, according to Reuters. First-quarter non-GAAP diluted EPS came in at 39 cents, up from 24 cents a year ago. The company still logged a GAAP net loss of $295.6 million.
Amazon gets a clear win for AWS as cloud giants try to grow custom silicon sales. Microsoft is pushing its Cobalt Arm chips for Azure, and Google Cloud is selling Axion, its custom Arm-based chips. Large providers are looking for ways to cut costs and lower power use.
Snowflake reported passing $7 billion in lifetime AWS Marketplace sales, and it now sees over $2 billion in calendar 2025 sales through that route. Customers like Fetch and Hex are running AI workloads on Snowflake using AWS with governed data, the company said.
AI demand could still stay lumpy. Snowflake’s product revenue only comes in when customers use its compute, storage or data transfer, not just when deals are signed. The company has said that remaining performance obligations aren’t a solid guide for future product revenue since the timing on usage can shift.
Snowflake competes with Databricks and the major cloud providers, though it partners with AWS. The new deal locks in capacity and deeper product ties. But it increases pressure—investors now want to see real, lasting usage, not just initial AI pilots that never go to production.
