Technological Overview

SQream Technologies’ patent pending technology boosts analytics performance through Massive Parallel Computing using GPU technology (Graphic Processing Unit). This revolutionary technology enables processing and analyzing of big data significantly faster than leading DBMSs and analytics solutions on the market today.

SQream’s analytics engine performance, surpasses database analytics by orders of magnitude, and is capable of processing and analyzing high volumes of data, while delivering a high cost/performance ratio.

With SQream the power of a full-rack database machine is condensed into a single standard server, delivering a superior cost-effective performance for Big Data.

SQream DB transforms the way organizations deal with their big data, through translating relational queries (SQL) into highly parallel algorithms. The process is executed efficiently through the use of GPUs, resulting in a unique high-speed, real-time, high scale performance.

SQream DB enables rapid implementation: no data modeling is needed, no new DBA skills are required, no new and expensive development and usage skills need to be learned/acquired, and up to 100TB of raw data can be stored and queried in a standard 2U server.

100 Times Faster Insights


SQream Technologies provides you with best of breed big data features enhanced with GPU technology, providing up to 100 times faster insights from data. The database is designed ground up to support columnar and SPMD technology (Single Program, Multiple Data).


  • Inserts and analyzes hundreds of Billions of records in seconds
  • Massive parallel computing power
  • Faster compute – X20 more processing power in each node
  • Aggressive data crunching – X50-70
  • Faster, higher scalability – scaling in GPU cards, not in servers
  • 50% less time to implement
  • 90% less footprint
  • 90% less power consumption



SQream delivers massive performance in a box:

  • Massive GPU processing power
  • Aggressive compression resulting in a small footprint

Quick hassle-free implementation:

  • No indexes, no materialized views or cubes, just raw data
  • No change of code, no change of data science paradigms, same visual tools

Multiple architectures:

  • Standard servers (leading vendors), scalable cloud
  • Non-intrusive topology