Home > Workload Solutions > Data Analytics > Guides > Reference Architecture—Multicloud Data Analytics with Dell Technologies Powered by Starburst > Multicloud data analytics with Dell Technologies Powered by Starburst
This reference architecture showcases how you can simplify and accelerate the data analytics experience for data users by enabling a single point of access for all your data. It aims to enable immediate access to data across multiple database and storage systems, independent of whether they are on premises or in the public cloud.
Our architecture combines Dell PowerEdge servers and an ECS object storage functioning as data lake storage, together with the analytics query engine from Starburst. The Starburst Enterprise Platform builds on the popular Trino open-source SQL query engine, developed originally under the name of Presto at Meta. Trino queries can process data across multiple data systems, including storage systems containing data in open file formats such as Apache Parquet and open table formats such as Apache Iceberg that adds table structure to formats such as Parquet.
With a query engine such as Trino, we implement a layer on top of data that abstracts away details on location, connectivity, language variations and even API. Such a layer of abstraction is critical to simplify data analytics over a diverse set of data sources, including storage systems such as the Dell Elastic Cloud Storage, capable of storing large amounts of data using open formats.
We built the environment described in this document using Dell Technologies infrastructure and evaluated the performance of the Starburst engine using TPC-DS, a benchmark widely accepted by the industry. We have detailed our set up and experiments for reproducibility in this paper.