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

Source: TechCrunch

Snowflake has signed a new $6 billion five‑year agreement with Amazon Web Services, announced on Wednesday.

Deal Overview

• Snowflake has always run on AWS, though it is also available on Microsoft Azure and Google Cloud.
• According to AWS, Snowflake has sold $7 billion worth of services via the AWS Marketplace since its founding in 2012, making this new contract roughly equivalent to the total revenue it has ever generated from that cloud.
• Snowflake expects its spending on AWS to double to $2 billion in calendar year 2025.

Snowflake’s AI Offering

Snowflake’s AI building tool, Cortex AI, has been available for a couple of years. It leverages the fact that much of an enterprise’s data resides in Snowflake and provides features such as:

• Natural‑language text interface for database queries
• Automated summary reports

Focus on AWS Graviton CPUs

The contract includes expanded access to AWS’s home‑grown ARM‑based CPU, Graviton. As AI moves from training to daily usage and automation via agents, CPU demand surges. While GPUs handle training and inference, CPUs manage the majority of ancillary AI tasks, especially those involving agents.

• Amazon CEO Andy Jassy recently highlighted that AWS’s own AI chips offer “better price‑performance” than Nvidia’s, even though Nvidia GPUs remain a core part of the AWS offering.
• High demand for AI processing is prompting cloud providers to deploy chips at rapid pace.

Competitive Landscape

• Meta: Last month AWS signed a deal to supply millions of Graviton chips to Meta for its AI compute needs, following Meta’s earlier $10 billion agreement with Google Cloud.
• Google: Continues developing its own AI chips.
• Microsoft: Launched the Maia AI chip in January.

These moves signal growing competition for Nvidia’s dominance in AI hardware.

Nvidia’s Response

Nvidia CEO Jensen Huang announced the company’s new AI‑specific CPU, Vera, describing it as opening a “brand‑new $200 billion market for Nvidia.” Huang noted that Nvidia has already sold $20 billion of the product and emphasized the company’s readiness to defend and expand its market position.

Implications

While Nvidia remains a major player, the influx of multi‑billion‑dollar cloud deals involving alternative CPUs demonstrates how AI is expanding opportunities for cloud providers. As AI continues to integrate into work and home environments, the cloud ecosystem is poised to capture a significant share of the resulting value.