Snowflake Soars 33% After Earnings Beat and $6 Billion AWS Commitment

CNBC reported Wednesday that Snowflake shares surged as much as 33% in after-hours trading. The rally followed a strong first-quarter earnings print and a landmark Snowflake AWS deal worth $6 billion over five years.

A Blowout Quarter Fuels the Rally

Snowflake posted adjusted earnings of 39 cents per share on $1.39 billion in revenue. That represented 33% year-over-year top-line growth. Analysts polled by LSEG had penciled in 32 cents per share and $1.32 billion in revenue. Forward guidance also exceeded expectations. Snowflake projected second-quarter product revenue of $1.415 billion to $1.420 billion. Analysts had anticipated roughly $1.37 billion. The company also flagged an expected adjusted operating margin of 12.5%, above the 11.9% consensus estimate.

The Amazon Deal and Graviton Push

Amazon Web Services confirmed the five-year spending arrangement in a press release. The Snowflake AWS deal averages roughly $1.2 billion annually. It includes expanded use of Amazon’s Graviton chips, the company’s custom Arm-based processors, alongside cloud-based GPUs for AI workloads. Amazon first launched Graviton in 2018. Snowflake publicly discussed adopting the architecture as far back as 2022. The agreement contains no equity component, distinguishing it from Amazon’s deals with Anthropic and OpenAI.

Background: A Long Partnership Deepens

Snowflake has relied on AWS since going public in 2020 with a market cap that now sits just above $60 billion. At IPO, the company disclosed a $1.2 billion cloud-provider commitment over five years. That provider was later confirmed to be Amazon. By 2023, the arrangement had grown to $2.5 billion. The new $6 billion figure marks a dramatic acceleration in that relationship. Snowflake also announced plans to acquire AI startup Natoma for an undisclosed price, signaling continued investment in artificial intelligence capabilities.

Why Graviton Matters for Agentic AI

The chip architecture choice reflects a broader market shift. For decades, server infrastructure ran on x86 processors pioneered by Intel and later AMD. Arm-based designs, which prioritize power efficiency, entered data centers through Amazon before spreading to Google and Microsoft. In the current AI cycle, demand for general-purpose compute is rising alongside the growth of agentic applications. Unlike chatbots, agentic AI systems perform multi-step tasks and move large volumes of data continuously. That workload favors CPUs like Graviton over pure GPU setups. Snowflake also maintains a separate partnership with Nvidia focused on running AI workloads on GPU infrastructure, announced in 2023.

Read Next: AWS Confirms Anthropic Will Spend Over $100 Billion on Cloud Infrastructure