In more good news for Amazon, Snowflake signs $6B deal with AWS for AI CPU chips

Cloud data platform Snowflake has entered into a new five-year agreement worth $6 billion with Amazon Web Services (AWS), highlighting the growing role artificial intelligence is playing in driving enterprise cloud spending.

The deal marks one of the largest commitments between the two companies to date. Snowflake has historically operated on AWS infrastructure, although its services are now also available through Microsoft Azure and Google Cloud. To put the size of the agreement into perspective, AWS noted that Snowflake has generated approximately $7 billion in sales through AWS Marketplace since the company's founding in 2012. The new contract alone approaches that lifetime figure.

According to Snowflake, customer demand has been a major factor behind the expanded agreement. The company said spending by its customers on AWS services accelerated significantly in 2025, reaching roughly $2 billion, more than double previous levels.

Artificial intelligence remains a key driver of this growth. Snowflake has spent the last several years expanding its AI capabilities through Cortex AI, a platform designed to help businesses work more effectively with the large volumes of data stored within Snowflake environments. The technology allows users to interact with databases using natural language, generate summaries, and perform advanced analytics without requiring traditional coding expertise.

A notable aspect of the new agreement is Snowflake's increased use of AWS's custom-built Graviton processors. These ARM-based chips have become increasingly important as AI workloads evolve beyond model training and move into large-scale deployment and automation.

While graphics processing units (GPUs) remain essential for training and running advanced AI models, central processing units (CPUs) handle many of the supporting tasks required for AI applications, particularly autonomous agents and enterprise workloads. As organisations adopt AI at scale, demand for processing power across both CPUs and GPUs continues to rise.

Amazon has been aggressively promoting its in-house chip technology as a lower-cost alternative for cloud customers. AWS CEO Andy Jassy recently highlighted the price-performance advantages of Amazon's custom chips, while continuing to support Nvidia hardware throughout the company's cloud infrastructure.

The strategy appears to be attracting major customers. Earlier this year, AWS secured a significant agreement with Meta to provide millions of Graviton processors for the company's expanding AI operations. That deal followed Meta's separate multibillion-dollar cloud agreement with Google Cloud, underscoring the intense competition among major cloud providers for AI-related business.

The broader trend also reflects growing competition in the semiconductor market. Alongside Amazon, companies such as Google and Microsoft have invested heavily in developing their own AI-focused processors. Microsoft introduced its Maia AI chip earlier this year, while Google has continued expanding its custom chip initiatives.

Despite these efforts, Nvidia remains a dominant force in AI infrastructure. CEO Jensen Huang recently emphasised the company's ambitions to expand its market position, citing Nvidia's newly launched AI-focused CPU platform, Vera. Huang described the segment as a massive opportunity and said demand for the company's latest products remains exceptionally strong.

Although cloud providers are increasingly promoting their own chip technologies, Nvidia continues to play a central role in AI computing. Most leading AI models and applications are still heavily optimised for Nvidia hardware, making the company a critical supplier across the industry.

For AWS, however, the latest Snowflake agreement demonstrates how the AI boom is translating directly into large-scale cloud contracts. As enterprises increase their investment in artificial intelligence, cloud providers are positioned to benefit regardless of which AI platforms or applications ultimately dominate the market.

The growing demand for computing power, data processing, and AI infrastructure is creating significant opportunities across the cloud ecosystem, and the new multibillion-dollar partnership between Snowflake and AWS is another example of how rapidly the market continues to expand.