Our GPU accelerated columnar database “qikkDB” is now available not only on Amazon AWS, on DockerHub but also under the Apache 2.0 open-source license on GitHub! In open-sourcing qikkDB we have made a step not just to bolster the community around qikkDB, but also to further the global innovation cause.
What is qikkDB?
qikkDB is an ultra fast database that uses graphic cards (GPUs) to accelerate computations over big data. In Instarea we work with big telco data what makes us challenged everyday with optimal approaches to manipulation with massive datasets. This project started as a reaction to our needs for fast querying over long tables filled with event records.
The main focus of qikkDB is to query a single flat and huge table with realtime response. This approach will find its place in use-cases in which speed is crucial e.g. fraud detection in finance data, web logs analysis and realtime decisions, realtime population analytics on telco data, etc.
qikkDB is already suitable for
- Filtering and aggregations over single flat huge table
- Complex polygon operations (contains, intersect, union)
- Numeric and datetime data
- Incremental data which are growing over time
Not yet suitable for
- Joining normalized data in multiple tables
- Complex string manipulation
- Modification of historical data
- Advanced scripting (procedures)
Engine for Advanced Visualization
qikkDB is seamlessly integrated to the visualization tool TellStory. This tool enables the creation of visual stories over big data with realtime responses. To learn more about TellStory see visit this page.