A startup launched by HPC veterans is combining supercomputers and graph analytics to forge a new suite of tools for applications ranging from cyber security to scientific research.
Along with the launch of its xGT graph analytics tool, Seattle-based Trovares announced the close of a Series A funding round. The funds will be used to expand the new platform’s features, the startup said Monday (April 15).
The property graph startup identified the Defense Department as an early customer for its xGT platform, which it is positioning as a highly parallel graph analytics tool capable of squeezing full utilization out of every available processor. The tool also uses a framework called “symmetric multiprocessor systems” (SMP) to perform trillions of scans of in-memory graphs. That approach helps spot complex patterns, often in minutes, the startup claims.
The xGT graph analytics tools “was designed to scale to very large data sets that other vendors cannot address,” said Trovares CEO James Rottsolk, who previously served as chief executive of supercomputing pioneer Cray Inc. “While conventional graph analytics tools post benchmarks up to two times faster than conventional tools, Trovares claims xGT can deliver speed and performance that are “orders of magnitude faster and larger than conventional tools.” (The company offered the benchmarks show below but didn’t identify the competitor or system specs, click to enlarge.)
Graph analytics databases store data in a graph format that includes nodes, edges and properties, making them better suited to handling large and complex data sets. Trovares claims the SMP framework also allows its HPC-powered tool to traverse “billion-edge” in-memory graphs to spot complex patterns.
Trovares joins a growing list of graph database vendors looking to deliver faster queries and improved analytics. Along with spotting cyber intrusions and detecting “fraud signals,” the startup is targeting a growing list of graph analytics applications ranging from scientific research to “social network influencers.”
The startup touts its HPC background as a key market differentiator, noting that its xGT platform scales to handle very large data sets.
Rottsolk headed Cray for 18 years. Co-Founder David Haglin, who serves as CTO of Trovares, previously worked as a research scientist at Pacific Northwest National Lab (PNNL). Richard Russell, the startup’s director of technical marketing, also worked at Cray before joining PNNL.
The startup said its xGT graph analytics tool is available now for on-premise or cloud deployments, including on Amazon Web Services (NASDAQ: AMZN) for data sets up to 4 terabytes. (AWS launched a managed graph database service in November 2017 called Neptune.) The xGT application also can run on existing IT infrastructure.
This article first appeared in HPCwire’s sister publication Datanami.