Snowflake Gen2 Warehouses: Blazing Fast Performance, Now Available on All 3 Clouds Across Expanded Regions

Since Snowflake’s founding in 2012, we have striven to be the market leader in fast performance for all workloads and continue to iterate rapidly to automatically improve by doing the heavy lifting on behalf of our customers. Our philosophy is simple: Our customers’ success is our success. 

In 2012, Snowflake revolutionized the data world with our virtual warehouses that had automated cluster management, maintenance, upgrades and regular performance improvements — all without any downtime — for a truly easy-to-use service. In 2023, we introduced Snowpark Optimized warehouses, which allowed data scientists and data engineers to run large and memory-intensive machine learning (ML) training jobs and native code that relies on custom packages via Snowpark. Most recently at Snowflake Summit in June, we announced the general availability of Standard Warehouse - Generation 2 (Gen2) (in limited clouds and regions), an updated version of Snowflake’s current Standard Warehouse (Gen1) that has upgraded hardware and additional performance enhancements. We are excited to announce that Gen2 is now generally available across all three clouds — AWS, Microsoft Azure and Google Cloud Platform (GCP) — across an expanded set of regions. 

Figure 1: Snowflake Virtual Warehouse Evolution
Figure 1: Snowflake virtual warehouse evolution diagram

With the upgraded hardware and additional performance improvements, Gen2 has delivered 2.1x faster performance for core analytics workloads on Snowflake tables over the last 12 months, ending May 2, 2025.

Figure 2: Gen2 delivers 2.1x faster analytics performance for Snowflake tables
Figure 2: Gen2 delivers 2.1x faster analytics performance for Snowflake tables.

Furthermore, customers  are seeing even bigger improvements on Gen2 for Data Manipulation Language (DML) (especially DELETE, UPDATE and MERGE) operations, data intensive analytics queries with large table scans and complex aggregations. In fact, as of May 2, 2025, Snowflake has delivered up to 4.4x faster performance for DELETE, UPDATE and MERGE DML operations, as compared to Standard Warehouses. Learn more about the DML innovations that have been incorporated into Gen2.

Figure 3: Gen2 delivers up to 4.4x faster performance for DELETE, UPDATE and MERGE DML operations.
Figure 3: Gen2 delivers up to 4.4x faster performance for DELETE, UPDATE and MERGE DML operations.

Hear from our customers and partners 

SELECT, a Snowflake partner that helps customers with advanced optimization and cost management, evaluated Gen2 across several workloads and found substantial point-in-time performance improvements of  46% for select queries and 49% for filtered DML, as compared to Standard Warehouses.

Customers like AppFolio and Babylist have seen significant performance improvements using Gen2:

Quote Icon

At AppFolio, our team is dedicated to delivering timely, organization-wide insights, a commitment upheld by our data refresh SLAs. As data volumes and sources rapidly expanded—each bringing unique complexities and data freshness requirements—maintaining these standards became increasingly challenging, often demanding intricate processes and occasional manual intervention. Now, with Gen 2 warehouses, we've achieved a significant performance boost. This not only allows us to consistently meet our SLAs with ease, but it also empowers our team to focus on generating greater strategic business value.

Ryan D'Souza
Staff Data Engineer, AppFolio
Quote Icon

“Our key question with Gen 2 warehouses was: Can we simply run our jobs faster? The answer is yes. We have a daily dbt project build with a 9 AM ET SLA that’s critical for downstream users, everyone from analytics and ML teams to our fulfillment centers. Since moving to Gen 2 warehouses, we’ve automatically cut daily runtime and now easily meet our SLA.”

Arthur Wang
Staff Data Engineer, Babylist

Additionally, some customers like Canva have not only seen tremendous performance improvements switching from Gen1 to Gen2 but have also saved on total costs by being able to better optimize their warehouses — enabling them to bring more workloads to Snowflake with the cost savings:

Quote Icon

Our product analytics event deduplication process is one of our most critical workloads at Canva — it runs around the clock to keep our data pipeline clean and accurate. When we upgraded the virtual warehouse from Gen1 to Gen2, we saw notably faster query performance, allowing us to ultimately optimize to save around 30% on costs. With these cost savings, we can now bring more workloads to Snowflake.

James Robinson
Senior Analytics Engineer, Canva

Get started with Gen2 warehouses today

You can easily create a Gen2 warehouse using the CREATE WAREHOUSE command or in Snowsight. 

Snowsight UI

In addition to creating new Gen2 warehouses, it is super easy to upgrade an existing standard Gen1 warehouse to Gen2. You can do it online, just by altering an existing warehouse, without any downtime. When you convert a Gen1 warehouse to Gen2, existing queries that were running on Gen1 continue to run to completion on Gen1 machines, while the new queries are executed on the Gen2 machines until the entire workload switches to running on Gen2.

CREATE OR REPLACE WAREHOUSE gen1_to_gen2_wh
  WAREHOUSE_SIZE = XSMALL;

ALTER WAREHOUSE gen1_to_gen2_wh
  SET RESOURCE_CONSTRAINT = STANDARD_GEN_2;

To learn more about Gen2, please refer to our documentation.

Author
Share Article

Subscribe to our blog newsletter

Get the best, coolest and latest delivered to your inbox each week

Where Data Does More

  • 30-day free trial
  • No credit card required
  • Cancel anytime