Snowflake Signs $6B AWS Deal to Scale AI with Custom Graviton Chips

Cloud data storage giant Snowflake has signed a massive five-year, $6 billion agreement with Amazon Web Services (AWS). The deal, announced on Wednesday, underscores a significant commitment to AWS infrastructure as Snowflake looks to scale its AI capabilities. The scale of the contract is particularly notable; for comparison, Snowflake has sold $7 billion worth of its services via the AWS Marketplace in total since the company was founded in 2012.

A Strategic Shift Toward Graviton Chips

The core of this multi-billion-dollar agreement centers on Snowflake’s increased access to AWS’s homegrown, ARM-based CPU chip, known as Graviton. While GPUs are primarily utilized for AI training and reasoning, CPUs are essential for the daily usage and automation tasks performed by AI agents. As AI adoption shifts from initial training to active, large-scale deployment, the demand for CPU processing power has skyrocketed.
Amazon has positioned its custom silicon as a more affordable alternative to Nvidia’s offerings, claiming better price-performance. By deploying these chips, Amazon aims to pass cost savings along to its customers. This strategy has already attracted other major players, including Meta, which recently signed a deal to utilize millions of Graviton chips for its own AI compute needs.

The Growing Competition in AI Hardware

The rise of custom silicon from cloud providers like Amazon, Google, and Microsoft represents a direct challenge to Nvidia’s dominance in the AI hardware market. While major AI model makers have historically architected their applications specifically for Nvidia’s chips, the industry is seeing a shift as cloud giants invest heavily in their own proprietary hardware. Microsoft launched its Maia AI chip in January, and Google has been developing its own AI chips for years.
Nvidia, however, is not yielding its market position easily. CEO Jensen Huang recently announced a new AI-specific CPU called Vera, describing it as a $200 billion market opportunity for the company. Huang noted that Nvidia has already secured $20 billion in sales for the new technology, signaling a robust defense of its turf as cloud providers continue to scale their own internal infrastructure.

AI Driving Cloud Growth

Snowflake’s decision to deepen its reliance on AWS is driven by the rapid acceleration of its customers' spending on the platform. Snowflake reported that customer spending on AWS doubled in 2025, reaching $2 billion for that calendar year alone. Much of this growth is attributed to the adoption of Cortex AI, Snowflake’s tool that allows enterprises to interact with their data using natural language, generate summary reports, and perform complex database queries.
As AI continues to integrate into daily work and home life, the cloud providers are positioned to capture significant value. Whether through the deployment of custom chips or the hosting of advanced AI tools, the multi-billion-dollar deals between industry giants highlight how the rise of artificial intelligence is fundamentally lifting the business of cloud computing.