Your AI workloads don't follow a predictable pattern. Dashboards spike, agent activity fluctuates, and modern AI apps introduce unpredictable bursts of data prep, retrieval and inference requests. Your compute has to be agile enough to respond to those changes without requiring constant tuning, no matter where your workloads run.
Snowflake Adaptive Compute automatically adjusts compute resources to address fluctuating workload demands, reducing the need for time-consuming manual tuning and maintenance. Newly released enhancements are designed to deliver improved real-world price-performance for variable workloads, and those benefits are now generally available to customers on Microsoft Azure and Google Cloud (joining the previously-announced AWS availability).
A compute engine built for diverse AI workloads #
Instead of forcing you to choose between overprovisioning for peak traffic or risking slowdowns, Adaptive Compute is designed to continuously adjust compute resources in response to changing workload demands. This helps maintain performance while reducing manual tuning, so your team can focus on delivering insights instead of managing warehouses. With today's launch expanding Adaptive Compute to select regions across AWS, Microsoft Azure and Google Cloud, you can extend these adaptive benefits across your multi-cloud environment.
Improved price-performance for variable workloads #
Our latest improvements to Adaptive Compute deliver real-world price-performance gains for lower-concurrency and variable compute profile workloads, where demand fluctuates across the day or spikes unpredictably.
Compared to the Adaptive Compute version released in June 2026, customers may now see up to a 30% improvement in cost depending on the shape of the workload. That means Adaptive Compute is now an even more compelling choice for organizations with multiple types of demanding and dynamic workloads.
Now available in select Azure and Google Cloud regions #
Adaptive Compute is now generally available for Enterprise+ Accounts in select Microsoft Azure and Google Cloud regions, as well as the AWS regions announced in June:
Microsoft Azure
Americas: Central US (Iowa), East US 2 (Virginia)Europe: West Europe (Netherlands), North Europe (Ireland), UK South (London), Switzerland North (Zurich), Sweden Central (Gävle)
Google Cloud
Americas: us-east4 (Northern Virginia)Europe: europe-west 2 (London), europe-west 3 (Frankfurt), europe-west 4 (Netherlands)Asia Pacific: Australia-Southeast 2 (Melbourne)
AWS
Americas: US West 2 (Oregon), US East 2 (Ohio)Europe: EU West 1 (Ireland), EU Central 1 (Frankfurt)Asia Pacific: AP Northeast 1 (Tokyo), AP Southeast 2 (Sydney)
How to get started with Adaptive Compute #
If you're already using Snowflake, moving to Adaptive Compute is seamless. Instead of rebuilding your infrastructure or changing warehouse names in your existing pipelines and scripts, you can upgrade your existing Snowflake standard warehouses with a simple configuration change and zero downtime. This allows you to transition your production and AI workloads smoothly while fully retaining your existing configurations, access control policies and monitoring metrics.
alter warehouse <warehouse_name> set warehouse_type='ADAPTIVE';
For setup guidance and benchmark results across analytics, throughput and DML workloads, check out our June launch post.
Next steps #
Learn more about Snowflake’s compute enhancements and how your organization can benefit from better performance with less operational complexity:
Watch our
Snowflake AI Pulsewebinar and see how Adaptive Compute sets up AI for successExplore
Adaptive Computecapabilities and use casesGet more details about availability, requirements and setup in our
Forward-looking statements: This content contains forward-looking statements, including about our future product offerings, and are not commitments to deliver any product offerings. Actual results and offerings may differ and are subject to known and unknown risk and uncertainties. See our latest 10-Q for more information.