SQreamDB

A SQL database that empowers organizations to perform complex analytics on petabyte-scale of data and gain time-sensitive business insights, faster and at one tenth of the cost.

 

Asking Bigger with SQreamDB

A SQL database that empowers organizations to perform complex analytics on petabyte-scale of data and gain time-sensitive business insights, faster and at one-tenth of the cost.

Data Warehouse

Store and manage enterprise-scale data, so decision-makers, business analysts, data engineers, and data scientists can analyze the data and gain valuable insights from BI, SQL clients, and other analytics apps.

SQreamdb data warehouse

  • 500TB workload, 7 GPU’s
  • Ingestion rate: ~9TB/Hour
  • 10 hours of 20 complex queries

Data Preparation

Transform raw data through denormalization, pre-aggregation, feature generation, cleaning, and BI processes, so it can be ready for Machine Learning and AI processes.

SQreamdb data preparation

  • Ingestion rate: ~5TB/Hour per GPU
  • Shortening data preparation processes by 50-80%

Query Engine

Analyze data from any source, in any technology, and in any format, on top of existing analytical solutions and without any data duplication required.

SQreamDB query engine

  • 50+ sources and formats
  • Supporting all Data lakes

Ecosystem

businessIco Business Intelligence and Visualization
dig deeper ML & Data Science
dataIco Data Integration

Product highlights

query

Query

Processing is performed by using the ANSI-SQL syntax. Running queries can be done through the built-in SQream Acceleration Studio, or through a third-party BI tool.

connectivity

Connectivity

SQream integrates into existing ecosystems, with support for industry-standard ODBC and JDBC connectors, as well as Python and C# .Net, C++, Java, and others.

parraleism

Parallelism

The GPU is used to achieve parallel data processing. By splitting large tasks into smaller processes, SQream distributes operations between multiple GPU cores.

compression

Compression

All the data that is it ingested is automatically compressed at a 5:1 ratio

architecture

Architecture

The compute and storage are completely separated, with multiple compute units, running to store or retrieve data from a single or multiple storage sources. This concept provides flexibility and easy scaling, while data processing is being done not in memory.

performance

Performance

Acceleration leans on synchronizing all available resources (CPU, GPU, RAM) for complex analytical tasks while performing automatic vertical and horizontal partitioning of the data. moreover, it stores data tables by columns, therefore eliminating unnecessary reading for each analytical workload.

Start Asking Bigger

Find out why organizations ranging from fast-growth startups to Fortune 100s all rely on SQream.