Snowflake expanded its Data Cloud with Unistore Hybrid Tables that enable organizations to optimize data for both analytics and transactional workloads.
The cloud data platform vendor introduced the new features on Tuesday at its Snowflake Summit 2022 conference.
Initially solely a cloud data warehouse, Snowflake in recent years has expanded into a cloud data platform that enables users to build applications with Snowpark and handle data lake workloads and unstructured data sources.
At Snowflake Summit, the vendor, based in San Mateo, Calif., said it is also working on support for Apache Iceberg data lake tables, though there isn't yet a public preview.
Iceberg is an increasingly popular data lake technology that is being supported by a growing number of data vendors, including Dremio and Starburst Enterprise with its data lakehouse engine.
But while Iceberg is a popular data lake format, it isn't the only one. Snowflake does not yet support the popular open source Delta Lake format, which was originally created by AI and data lake vendor Databricks.
Snowflake also introduced in preview its new Unistore capability, a hybrid table that will enable users to run both analytics and transactional workloads.
The ability to run both kinds of workloads is a fast-growing growing trend, with several data vendors in recent months coming out with similar services, including Oracle's MySQL Heatwave, Google's AlloyDB database and new updates that are coming to the MongoDB NoSQL platform.
The Unistore capability is of particular interest to a number of Snowflake users, who detailed their expectations and initial applications during a media roundtable before the Snowflake summit.
A group at software giant Adobe uses Snowflake for a number of applications, including to support Adobe Campaign, a cross-channel campaign management application.
Snowflake helps Adobe scale Adobe Campaign workloads so users can run their campaigns at larger scale and with shorter timelines, said Sunil Menon, head of product marketing and go-to-market strategy for the B2C customer journeys portfolio at Adobe.
"The typical campaign execution process starts with identifying the right customers to target and then building a very personalized message," Menon said. "For every individual customer, that translates into query and analytical workloads that really need to scale as our customers bring more and more data into the platform."
Snowflake expands data cloud with Unistore Hybrid Tables
In addition to analytics workloads, Menon noted that Adobe also runs transactional workloads such as sending out order confirmation emails and enabling users to update preferences.
To date, Adobe has had to use two systems for its data architecture: one optimized for transactional workloads, and the other for analytical workloads.
Carl PerryDirector of product management, Snowflake
With Unistore and Hybrid Tables, Menon said Adobe will be able to simplify its data architecture with Snowflake by using a single architecture for both types of workloads.
"Unisource is really a modern approach to working with both transactional and analytical data in a single unified platform," said Carl Perry, director of product management at Snowflake.
How Iceberg expands Snowflake for data lakes
With Apache Iceberg, Snowflake is looking to ramp up its ability to support data lakes.
The vendor's goal is to fully support Iceberg tables with the ability to insert, update, delete and create new data tables, said Christian Kleinerman, senior vice president of product at Snowflake.
Kleinerman said some Snowflake customers have been asking for open file formats for data lakes to improve interoperability.
"We're doing this as a matter of choice," he said. "Our customers and our partners will be able on a table-by-table basis to choose what should be in open table formats versus the existing Snowflake format."