A database designed from the ground up
to deliver truly real-time insights

GPU acceleration changes the game for big data analytics

GPU Processor

GPUs are designed around thousands of small, efficient cores that are well suited to performing repeated similar instructions in parallel. This makes them well-suited to the compute-intensive workloads required of large data sets.
Some of the latest high performance GPUs from NVIDIA feature over 4,000 cores, versus 16 to 32 cores per typical CPU-based device.
GPU cores crunch data far more efficiently and quickly than CPUs—which process data sequentially.

So much raw compute power, you won't need to rely on indexes and partitions


Very Low Latency

With less to index, data is available immediately after it is written. No more waiting for data structures to be updated before the data can be returned in queries.


Ask any question

Since there is no need to prepare the schema before it can be explored, business analysts have complete flexibility and freedom for data discovery projects.


Linear Scale out

With simpler data structures, the database scales in proportion to the size of the data. This leads to smaller and more predictable hardware costs.

Did we mention speed?

10x-100x faster than even the most advanced in-memory databases

Incredibly Fast Ingest

Each node in the cluster can ingest data directly without management from a head node. The result? Incredibly fast ingest.

Blazing fast query Response

In tests, Kinetica returns results for advanced analytical queries on billions of rows of data in well under a second.HANA

GPU benchmarks Legend
Link off to an article on benchmarks

Scale out on industry-standard hardware

The efficiencies of Kinetica's GPU architecture typically result in hardware costs that are 110 that of standard in-memory databases.


Kinetica runs on commodity hardware, that scales out linearly to meet your needs.
Typical hardware setup: 256GB - 1TB memory with 2-4 GPUs per node.
Kinetica's consolidated footprint requires a fraction of the power and cooling of traditional in-memory systems.

Certified to run on premise with:


Or in the cloud:

Accelerated By
NVIDIA partnered with Kinetica

Set Business Users Free!

Kinetica provides a range of methods for analysts and line of business users to query across the full dataset without the need for partitioning, downsampling, or redesigning the data model when a new query capability is needed.


Full Text Search

Full text search functionality is supported by Natural Language Processing (NLP) and the full set of results can be rendered on a map. Results aren’t limited to the first 10, 20 or 30 pages of relevant information.


Write Queries in SQL

Kinetica is a fully SQL compliant database; no programming necessary. Query your data the way you want to.


Use with Popular BI Tools

ODBC/JDBC connectors ensure data can be used with popular BI tools such as Tableau and open source front-end tooling such as Kibana and Caravel.

Native Visualization & Geospatial Tools

The GPU opens up incredible improvements for visual rendering of data—particularly geospatial and temporal data. Kinetica enables you to plot billions of data points and see changes in real-time as the underlying data or queries change.

  • Visualize billions of data points, and display changes in real time as the underlying data changes.
  • Take advantage of the full gamut of geospatial visualization renderers such as feature, class break, heat map,and geofencing.
  • Leverage Kinetica’s rich library of visualization object types, or connect to your favorite BI and/or geospatial mapping tool.
Vary icons above. New MBP in picture
Kinetica Visualization Demo
Kinetica is able to render billions of points on a map in under a second. Watch the demo »

Try out an interactive demo

Our demo, running on a 4 node cluster, is connected to a selection of live data feeds including Twitter, FAA flight data, shipping data and more.

Power and capability for developers and administrators

Comprehensive APIs

Kinetica can easily be interacted with via RESTful HTTP endpoints. Supports both JSON and Avro serialization. Open source native language bindings can be found for Java, Python, Javascript, and C++. More »

Ingest from multiple sources

Pre-built connectors for Apache Kafka, Apache NiFi, Apache Spark, Spark Streaming, Apache Storm, and ODBC/JDBC make it simple to ingest data from a wide range of data sources.
See Examples »

Scale out

Kinetica provides linear scalability on industry standard hardware. Data replication is handled for high availability. Sharding can be done automatically, or specified and optimized by the user.

Kinetica Architecture

Take the Kinetica Challenge!

Sometimes benchmarks and marketing copy can sound too good to be true. The best way to appreciate the possibilities that GPU acceleration brings to large-scale analytics is to try it with your own data, your own schemas and your own queries.

Contact us, and we'll set you up with a trial environment for you to experience it for yourself.