Kinetica debuts GPU-accelerated database for analyzing streaming data with improved performance, visualization

Kinetica released Wednesday lastest version of its distributed, in-memory database accelerated by GPUs that simultaneously ingests, explores, and visualizes streaming data.

Version 6.0 features advances in performance, visualization and high availability for enterprise-grade stability and security. Kinetica empowers organizations to perform a wide range of use cases in real time, including sentiment analysis, anomaly and fraud prevention, resource allocation on the fly, terrorist and other national security threat tracking, energy generation optimization, inventory tracking and customer engagement improvements.

Version 6.0 of Kinetica includes faster performance leading to a more optimized and tuned experience; and NVIDIA NVLink support to boost database performance as data moves between GPU and CPU three times faster on average compared to the traditional PCI Express. Its advanced visualization delivers 3D acceleration with GPUs to deliver up to five times faster visualizations; comes with high availability for failover, so that no single point of failure with new intra-cluster failover protection, including native cluster resiliency. Even if a node fails, database will keep working.

The release also delivers full SQL-92 query support to enable GPU-accelerated SQL-92 query support through certified JDBC and ODBC connectors. Its cluster resizing helps increase capacity of cluster more easily and more seamlessly by adding nodes while still online. It also delivers better cluster management for rebalancing workloads, and a new visual installer that allows for easy click button installation of Kinetica across hundreds of nodes.

Set to be available in the fourth quarter of this year, the Kinetica database makes “real time” a reality by ingesting large petabyte-scale, streaming data, while delivering analytic results and producing visualizations in milliseconds. Embedding GPUs into Kinetica’s architecture means there are 4,000-plus cores per device, versus 8 to 32 cores per CPU-based device. This translates into tangible savings from a smaller hardware footprint and less power and cooling.

By plugging into existing data architectures, Kinetica seamlessly delivers processing and analytics without the typical tuning, indexing, or tweaking associated with traditional CPU-based solutions. Data consumption is also simplified via free-text search, a native visualization engine, and plug-ins with third-party business intelligence applications such as Tableau, Kibana and Caravel.

Kinetica provides complete geospatial object track type support and a native geospatial visualization rendering engine that, in tandem with its parallel processing analytics capabilities, make it the best modern geospatial information system (GIS) for use cases where time and location matter. It costs a fraction of legacy tools, can alleviate costly middleware, and delivers much faster performance.

“Customers can accelerate their digital business solutions from 10-100x using Kinetica’s in-memory database accelerated by the immense compute power of NVIDIA GPUs,” said Jim McHugh, vice president and general manager at Nvidia. “The performance and efficiency of Kinetica’s distributed solution greatly exceeds that of traditional in-memory databases, giving customers significantly faster time to insights, while reducing their infrastructure cost and footprint.”

Existing common data management infrastructures are comprised of data warehouses and data lakes. For companies looking to derive insights in real time from live data such as Internet of Things (IoT), gaps remain from systems designed to periodically load data in from transactional systems (OLTPs). Moving from CPU to GPU-powered databases allows businesses to actually ingest, explore, and analyze massive streaming datasets in real time, while simultaneously reducing their hardware footprint.

“Due to the rise in IoT, there’s a firehose of data streaming from every channel, device, and human interaction with a tremendous potential to act on that data,” said Nima Negahban, CTO and cofounder of Kinetica. “Kinetica uniquely addresses the new paradigm of data by accelerating the data center infrastructure with NVIDIA GPUs, merging the query needs of the traditional relational database developer with the scalability demands of the modern IoT-centric enterprise. Kinetica seamlessly plugs into existing data architectures and ingests, analyzes and visualizes in real time virtually all data sources via connectors including ODBC/JDBC, Apache Hadoop, Apache Kafka, Apache Spark, Apache NiFi, and others.”

IoT Innovator Newsletter

Get the latest updates and industry news in your inbox! Enter your email address and name below to be the first to know.