Snowflake commits – Snowflake rallied as much as 30% after reporting a strong fiscal first quarter, then announced a new $6 billion, five-year cloud spending agreement with Amazon Web Services that includes Snowflake’s use of Amazon custom silicon and chips for AI workloads. The
On Wednesday, Snowflake’s stock reaction came fast—strong results, then a deal that signals where its compute power is heading next. Shares rose as much as 30% in extended trading after the company reported strong results for its fiscal first quarter, which ended on April 30.
The earnings beat landed with clear numbers: Snowflake posted 39 cents in adjusted earnings per share on $1.39 billion in revenue, up 33% year over year. Analysts polled by LSEG had expected 32 cents per share and $1.32 billion in revenue.
Snowflake’s outlook also impressed investors. The company called for a 12.5% fiscal second-quarter adjusted operating margin on $1.415 billion to $1.420 billion in product revenue. Analysts surveyed by StreetAccount had been looking for an 11.9% margin, with $1.37 billion in product revenue.
Then the bigger swing arrived with the announcement that Amazon Web Services has landed a $6 billion spending commitment from Snowflake for its cloud division. The agreement—described in a press release—covers Snowflake’s purchase of services and technology from AWS over five years.
At the center of that spend is AWS custom silicon built for AI. Snowflake’s purchases include the use of Amazon’s custom silicon and chips for artificial intelligence. and the company intends to expand its use of Amazon’s Graviton general-purpose chips. Snowflake also plans to increase its use of cloud-based graphics processing units for AI.
The new $6 billion arrangement implies an average annual spend of $1.2 billion.
Snowflake, which went public in 2020, has long relied on AWS. That relationship wasn’t new—but the scale is. At the time of Snowflake’s IPO. the company disclosed an amended deal with an unnamed cloud provider for $1.2 billion in spending over five years. with $350 million coming in the final year. A Snowflake spokesperson later told CNBC that the provider was Amazon. In 2023, that agreement climbed to $2.5 billion.
The AWS commitment also fits into a broader push inside the cloud industry as customers chase more advanced AI infrastructure. The article said the deal is the latest sign of AWS gaining momentum in AI as clients turn to the market-leading cloud for more advanced technologies.
In April, Claude creator Anthropic said it aims to spend over $100 billion on AWS over a decade. Amazon also has a deal with OpenAI.
Both of Amazon’s agreements with the AI model companies include an equity investment, while the Snowflake deal does not.
Alongside the cloud spending announcement, Snowflake disclosed a separate move: it said it was acquiring AI startup Natoma for an undisclosed sum.
Snowflake’s stock reaction reflects how much investors are watching not just earnings, but the direction of its computing strategy—especially as companies shift away from training AI alone and toward broader, task-driven systems.
The deal further places Snowflake among major tech firms choosing custom Arm-based processors over chips based on traditional x86 architecture. For decades. server chips have been built on the x86 instruction sets pioneered by Intel in the 1970s and Advanced Micro Devices a couple of decades later. Arm’s alternative power-efficient architecture went mainstream when Apple adopted it for the first iPhone in 2007. and it was Amazon that brought Arm chips into data centers with Graviton. Cloud rivals Google and Microsoft followed Amazon in bringing out custom Arm chips.
Fast-forward to 2026. when central processing units like Graviton are seeing a renewed flurry of demand as mass AI adoption moves from call-and-answer chatbots to task-oriented agentic apps. While GPUs like Nvidia’s excel at training AI models because they have thousands of tiny cores narrowly focused on performing many operations simultaneously. CPUs have a smaller number of powerful cores running sequential general-purpose tasks. Agentic AI requires a lot of general compute power to move large amounts of data around for AI workflows. orchestrating across multiple agents.
In April, Meta said it would draw on hundreds of thousands of Graviton chips. Amazon CEO Andy Jassy said in the company’s April earnings call. “Graviton is our industry-leading CPU chip. which allows Meta to run the CPU-intensive workloads behind Agentic AI with the performance and efficiency they need at their scale.”.
The Snowflake plan also has its own computing roadmap. It comes with close ties to Nvidia after a partnership announced in 2023. In November, Snowflake touted updates to simplify the process of running AI workloads on Nvidia GPUs.
Behind all of it is a simple reality for Snowflake: the company’s reliance on AWS is already established, and the new five-year, $6 billion commitment deepens it—this time with an explicit push into Amazon’s Graviton and cloud-based graphics processing units for AI.
For now, the market is reading the message the same way Snowflake is pushing it: earnings momentum is strong, but the bigger story is how the company intends to scale its compute for the next wave of AI workloads.
Snowflake AWS Amazon Web Services Graviton AI chips Natoma acquisition extended trading adjusted earnings per share product revenue agentic AI Arm-based processors