What Snowflake’s $6B AWS Deal Really Represents
Snowflake’s $6 billion (approx. RM27.6 billion) commitment to Amazon Web Services is a long-term cloud infrastructure spending agreement in which a software company prepays for Graviton CPUs and AI compute accelerators to secure the resources needed for data and AI workloads at scale. Unlike traditional pay-as-you-go cloud consumption, this multi‑year strategic collaboration agreement pushes Snowflake closer to the hardware layer, tying its growth to specific AWS services. The deal covers Arm-based Graviton processors for general compute and GPU‑accelerated EC2 instances for AI model training and inference, with AWS highlighting Snowflake’s role as a long‑time customer that has steadily shifted workloads to Graviton. In parallel, Snowflake is expanding its AWS footprint to additional regions and deepening its presence in AWS Marketplace, where lifetime sales have surpassed USD 7 billion (approx. RM32.2 billion), underscoring how infrastructure and distribution are increasingly bundled.
From Data Warehouse to AI Platform Built on Graviton and GPUs
Snowflake’s AWS alignment is tightly linked to its AI pivot under CEO Sridhar Ramaswamy, who describes the company as “the platform for the AI era.” Cortex AI, Snowflake’s suite of AI tools, runs close to customers’ governed data to support text‑to‑SQL, summarization, sentiment analysis, entity extraction, and even an AI coding agent via Cortex Code. To make these features viable at scale, Snowflake plans to run and train its generative AI models using a combination of AWS GPUs and Graviton CPU cores. According to Amazon, Snowflake’s lifetime AWS Marketplace sales crossed USD 7 billion (approx. RM32.2 billion) and exceeded USD 2 billion (approx. RM9.2 billion) during the 2025 calendar year, reinforcing AWS as both infrastructure base and sales channel. By standardizing more workloads on cost‑efficient Graviton instances, Snowflake aims to lower the unit cost of its traditional data warehousing services and free up spending capacity for compute‑heavy AI training and inference.
AI Compute Economics Push Software Vendors Into Infrastructure
Snowflake’s AWS deal shows how AI compute accelerators are changing the economics of enterprise software. Historically, software vendors avoided heavy infrastructure ownership, preferring cloud as a flexible operating expense. AI is changing that formula. Training and serving complex models, especially agentic systems that plan, retrieve data, and call tools, require dependable access to high‑end GPUs and large CPU fleets. To secure this capacity, companies like Snowflake are turning infrastructure into committed spend on their own balance sheets. This blurs the line between software provider and infrastructure investor, merging gross margin management with procurement strategy. The upside: predictable access to compute and the ability to defend premium AI offerings. The downside: if AI demand falls short, those long‑term commitments can squeeze profitability. The Snowflake AWS deal makes that trade‑off visible, signaling that software vendor partnerships with hyperscalers now revolve as much around capacity guarantees as product integrations.
Tighter Cloud Alliances and the Future of AI Infrastructure Spending
For AWS, Snowflake’s commitment reinforces its role as a preferred home for enterprise AI builds, combining scale, security, and mature procurement processes. AWS reported 28% revenue growth in the first quarter of 2026, suggesting enterprise AI spending continues to flow to major hyperscalers even as competition rises. At the same time, Snowflake’s decision not to emphasize AWS‑specific chips like Trainium, and instead focus on GPU‑accelerated instances and Arm‑based Graviton CPUs, points to a careful balance: deep integration without full vendor lock‑in. This approach supports multi‑cloud ambitions while still betting heavily on one primary partner. As more data and AI platforms follow this pattern, cloud infrastructure spending will increasingly appear as large, forward‑looking commitments by software companies, not only by cloud providers themselves. Snowflake’s USD 6 billion (approx. RM27.6 billion) bet signals confidence that AI‑accelerated compute will generate enough durable demand to justify long‑term capital exposure.