Snowflake Interactive Analytics: Breaking Down a Few Major Updates

In December 2025, Snowflake announced the general availability of Snowflake Interactive Analytics on AWS. Just a few months later, we have gotten great feedback from customers and are excited to announce a few big ways in which interactive tables and interactive warehouses have become even more powerful.

Support in Azure and GCP

In addition to the existing AWS support, we are excited to announce that interactive tables and interactive warehouses are now supported on Microsoft Azure and Google Cloud Platform (GCP). This means you can now build sub-second dashboards and support large-scale user-facing applications regardless of your chosen major cloud provider. You get the same market-leading performance and ease of use across the entire cloud ecosystem.

Manual scaling support for interactive warehouses

You can now manually adjust the multi-cluster warehouse (MCW) configuration of an interactive warehouse using the standard ALTER WAREHOUSE mechanism (generally available), allowing you to scale cluster counts as needed based on workload demand. This provides greater operational flexibility for managing concurrency and performance during peak usage. An automated scaling solution, similar to what Snowflake supports in standard warehouses, is currently in development and will further simplify workload management for high-concurrency scenarios.

Auto-suspend and auto-resume

You can now configure auto-suspend and auto-resume on interactive warehouses (generally available). The minimum auto-suspend setting, unlike in standard warehouses, is 24 hours. Because suspending an interactive warehouse clears the existing cache, this high minimum enables Snowflake to automatically suspend the warehouse only if the idle time is significant enough to warrant clearing the entire cache.

Search Optimization

The Search Optimization Service has long been the first-line optimization on Snowflake for achieving sub-second performance on point lookup queries. If you add Search Optimization on interactive tables, you can further accelerate point lookup queries, even those on massive tables.

For example, if you needed to do highly selective filters on a specific set of UserId(s), you could 1) put this data into an interactive table, 2) add it to an interactive warehouse and 3) create a Search Optimization access path for equality filters on UserId. When you run a highly selective query, Snowflake automatically prunes micropartitions using Search Optimization. The remaining micropartitions, ideally small after pruning, are served mostly from the interactive warehouse’s cache, limiting the amount of data that needs to be read from remote storage. The outcome: consistent sub-second performance, even on massive analytical queries.

Below, we compare the performance on queries on a standard table versus an interactive table with Search Optimization. We used the STORE_SALES table in the TPC-DS data set.

Specifically, we did 100 executions of a query with the below shape:

SELECT SUM(SS_NET_PROFIT) AS TOTAL_PROFIT
FROM STORE_SALES
WHERE SS_SOLD_DATE_SK = 2451432 AND SS_CUSTOMER_SK = 38027651;

We clustered STORE_SALES by (SS_SOLD_DATE_SK, SS_CUSTOMER_SK) for both the standard table and interactive tables. Additionally, we added Search Optimization for equality filters on these columns in the interactive table version. This helped queries on the column SS_CUSTOMER_SK because its cardinality is 65 million.

The results showed a big improvement in both pruning and query time compared with standard tables:

Metric	Standard Warehouse	Interactive Warehouse with Search Optimization
Average query time	144 ms	113 ms
Average bytes scanned	46.8 MB	1.6 MB
Average micro-partitions scanned	51	1

Replication

For mission-critical applications, downtime is not an option. We’re excited to announce that, like standard tables, you can now include interactive tables in your primary replication and failover groups (in private preview). Replication for interactive tables and warehouses is currently in private preview; please reach out to your account team to request access.

Storage lifecycle policies

We’re excited to announce that storage lifecycle policies are now supported for interactive tables (generally available) and interactive tables with autorefresh (private preview). You can now automatically archive or expire specific rows based on conditions such as data age, helping optimize storage costs while maintaining your compliance and governance standards. Snowflake executes these policies daily using shared compute resources. With lifecycle policies, you can retain high-performance interactive analytics on fresh data while seamlessly managing older data in a cost-efficient and compliant manner.

Next steps

You can get started using Snowflake Interactive Analytics today. Snowflake interactive tables and interactive warehouses best practices are largely the same as those for standard tables and standard warehouses. If you’ve already used Snowflake, Snowflake Interactive should feel familiar, and getting started takes just minutes.