Tech
May 28, 2026
Cloud data storage giant Snowflake has signed a $6 billion five-year agreement with AWS, primarily …
The Lead: Snowflake's $6B AWS Deal Signals Major Shift in AI Chip Market
Cloud data storage giant Snowflake has signed a landmark $6 billion five-year agreement with Amazon Web Services, primarily for access to Amazon's proprietary Graviton AI chips. This deal not only represents nearly all the revenue Snowflake has generated through AWS since its founding in 2012 but also signals a significant challenge to Nvidia's dominance in the AI hardware market as cloud providers develop their own competitive solutions.
The Event Details: Snowflake's Strategic Bet on AWS Graviton Chips
Snowflake and AWS announced the substantial agreement on Wednesday, continuing a partnership that has existed since Snowflake's inception. While Snowflake has expanded to run on Microsoft Azure and Google Cloud in recent years, this new contract demonstrates a deepening commitment to AWS infrastructure.
Of particular significance is Snowflake's specific request for increased access to AWS's home-grown ARM-based CPU chip, Graviton. This focus on Amazon's proprietary chips rather than relying solely on industry-standard solutions marks a strategic shift in how cloud providers and their customers are approaching AI infrastructure.
The timing of this deal aligns with Snowflake's acceleration of AI offerings, particularly its Cortex AI tool which has been available for several years. This tool enables natural language database queries and automated report generation - features that require substantial CPU processing power as AI applications move from training to everyday use.
The Data Analysis: $6B Deal's Financial Impact on Cloud Market
The financial implications of this deal are substantial. To put the $6 billion agreement in perspective, Snowflake has generated approximately $7 billion worth of services through AWS Marketplace since the company was founded in 2012. This means the new five-year contract represents nearly all the revenue Snowflake has ever brought in through AWS.
Furthermore, Snowflake reports that its customers are rapidly increasing their spending on AWS, with projections showing this spending will double to $2 billion in 2025 alone. This accelerated growth trajectory demonstrates how AI-driven workloads are creating significant revenue opportunities for cloud providers.
The deal follows AWS's recent agreement to provide millions of Graviton chips to Meta, representing another multibillion-dollar commitment to Amazon's proprietary silicon. These substantial contracts highlight how cloud providers are leveraging their custom chips to secure long-term partnerships and differentiate themselves in an increasingly competitive market.
The Impact Analysis: Cloud Giants Challenge Nvidia's AI Hardware Dominance
This deal represents a significant challenge to Nvidia's near-monopoly on AI processing hardware. While GPUs have traditionally handled AI training and reasoning, CPUs are becoming increasingly important as AI applications evolve toward daily usage and automation through AI agents. This shift in computing requirements has created an opening for alternative chip architectures.
Amazon CEO Andy Jassy has previously touted the Graviton chips' "better price-performance" compared to Nvidia's offerings, though AWS still utilizes Nvidia's chips in its cloud infrastructure. The strategy appears to be leveraging cost advantages to attract major customers, with Amazon passing these savings along to its clients.
Snowflake's decision to prioritize AWS's Graviton chips is part of a broader industry trend where cloud providers are developing their own silicon solutions. Google has long produced its own AI chips, while Microsoft recently launched its Maia AI chip in January. These moves collectively represent a concerted effort by cloud giants to reduce dependency on external hardware providers and increase control over their infrastructure stacks.
Nvidia CEO Jensen Huang has responded to this competitive pressure by launching the company's AI-specific CPU called Vera, which he claims represents a "$200 billion market opportunity" with $20 billion already in sales. This response underscores the high stakes in the AI chip market and the strategic importance of controlling the hardware layer of the AI stack.
The Prediction: The Future of AI Infrastructure and Cloud Competition
Looking ahead, we can expect the competition between cloud providers and specialized chip manufacturers to intensify. As AI applications become more pervasive and computationally intensive, the demand for efficient, cost-effective processing solutions will continue to grow. This will likely drive further innovation in both custom silicon and optimized software stacks.
The success of Snowflake's deal with AWS suggests that customers are increasingly willing to commit to proprietary chip architectures when they offer compelling performance advantages and cost benefits. We may see more large-scale, long-term agreements between cloud providers and their customers as both parties seek to optimize their AI infrastructure investments.
For the broader market, this trend toward cloud provider-developed chips could lead to a more fragmented but also more competitive AI hardware landscape. While Nvidia currently holds a dominant position, the entry of major players like Amazon, Google, and Microsoft with their own silicon solutions could ultimately benefit customers through increased innovation and potentially lower costs.
As AI continues to transform industries, the companies that control both the software and hardware layers of the AI stack will be positioned to capture the most value. Snowflake's $6 billion deal with AWS is just one example of how this strategic positioning is playing out in the rapidly evolving AI ecosystem.