Snowflake commits $6bn to AWS over five years, with Graviton chips at the centre

The five-year commitment is 2.4x larger than Snowflake’s 2023 AWS deal and lands as shares jump 38% on a Q1 earnings beat. The Graviton component is the part that matters strategically.

Snowflake has signed a five-year, $6bn commitment to Amazon Web Services in what both companies are framing as the largest expansion of their 11-year relationship to date.

The agreement, announced on Tuesday, includes commitments to run on AWS Graviton, the cloud provider’s custom Arm-based CPU line, and to develop deeper product integrations for what Snowflake is now describing as “agentic enterprise” workloads.

The trajectory of Snowflake’s AWS spending is the part that frames the size of the announcement. The company committed $1.2bn at its 2020 IPO. That figure rose to $2.5bn in a 2023 renewal.

TNW City Coworking space - Where your best work happens

A workspace designed for growth, collaboration, and endless networking opportunities in the heart of tech.

Book a tour now

The new $6bn agreement is roughly five times the 2020 commitment and 2.4 times the 2023 one.

The expansion mirrors Snowflake’s own growth: the company posted Q1 fiscal-2027 earnings on Wednesday that beat estimates substantially, and the stock jumped roughly 38% on the combined earnings and AWS-deal news.

The Graviton component is the part worth pausing on. AWS Graviton, now in its fourth generation, is Amazon’s in-house Arm-server processor line, designed to replace x86 chips from Intel and AMD inside AWS data centres at substantially better price-performance.

Snowflake committing to run its data-cloud workloads on Graviton at scale is a meaningful endorsement of the Arm-server thesis that has been quietly reshaping cloud-infrastructure economics for five years.

It is also a useful Amazon-side data point against the backdrop of yesterday’s news that ByteDance is building its own Arm and RISC-V CPUs to escape Intel and AMD pricing pressure.

The migration to custom Arm-server silicon, hyperscaler-led, is now visibly the structural story in data-centre CPUs.

For AWS specifically, the Snowflake deal lands inside a stretch of large AI-infrastructure commitments. Anthropic has committed to large AWS spending over multi-year terms; OpenAI signed a meaningful Microsoft Azure-competitor agreement on AWS earlier this year; Meta has been visibly expanding its AWS footprint for inference workloads. Snowflake is the latest in that sequence and the largest non-foundation-model commitment on the list.

AWS’s capacity to absorb $6bn of additional five-year demand against an already-stretched data-centre pipeline is itself a useful indicator of how rapidly Amazon is bringing new capacity online.

The strategic context for Snowflake is the agentic-AI thesis the company has been visibly betting on.

Snowflake’s pitch, like that of every enterprise-data-platform vendor at scale right now, is that AI agents will operate primarily over the trusted, governed enterprise data already inside customers’ cloud-warehouse environments, rather than over external training corpora.

Building that future requires substantially more compute integration with the underlying cloud provider; it requires direct access to AWS’s native AI primitives (Bedrock, SageMaker, the Q assistant) and deeper marketplace and go-to-market integration.

The remaining open question is the Databricks comparison. Snowflake’s most visible competitor is more directly bundled with Azure (through its 2023 Microsoft partnership) and has been positioning aggressively on multi-cloud agnosticism.

Snowflake’s deeper AWS commitment, including the explicit Graviton anchor, signals a different strategic bet: pick the larger hyperscaler partner, lock in customer-acquisition flow through AWS Marketplace, and accept the implicit single-cloud tilt that comes with it.

Whether that pays off against Databricks’s diversification posture is the multi-year question.

AWS Marketplace sales for Snowflake doubled year-on-year to $2bn in 2025, which suggests the integration logic is already working commercially.

Neither company disclosed which specific Graviton generation Snowflake is committing to. Snowflake CEO Sridhar Ramaswamy said the company will publish more detail at the AWS re:Invent conference later this year.

Story by Alina Maria Stan

Alina Maria Stan builds connections that people actually feel. As co-founder and COO of Tekpon, she turns product intuition into real moment (show all) Alina Maria Stan builds connections that people actually feel. As co-founder and COO of Tekpon, she turns product intuition into real moments of discovery, shaping how teams find and adopt SaaS every day. Since 2020, she has led Tekpon’s brand voice, media strategy, and growth plays with a clear focus on human outcomes behind every metric. Before Tekpon, Alina followed curiosity across industries and countries. She was CEO of King Casino Bonus and led affiliate and brand strategy at Extremoo Media and Fable Media in Denmark, where she learned how to build partnerships that last. Early on, she sharpened her CRM and pricing instincts at K.H. ApS, always asking why customers choose what they choose. Her approach is rooted in more than a decade of international experience and two master’s degrees, one in Sustainable Consumption from the Technical University of Munich and one in Consumer Affairs Management from Aarhus University.