Kinetica Db Inc. today said it has closed a $50 million series A financing round, nearly quintupling its total venture capital investment in one fell swoop.

The in-memory relational database maker, which is distinctive for its use of graphics processing units to accelerate performance, said it will use the funds to invest in engineering, sales, marketing and developing a cloud version of its product.

Kinetica’s engine addresses one of the major shortcomings of relational databases in general: difficulty processing large amounts of streaming data. It claims its converged platform can scale up to handle billions of records in data tables for online analytical processing, location-based analytics and machine learning applications.

GPUs are a key factor in the equation. Originally developed for applications such as visualization and gaming, GPUs are increasingly finding use as parallel processing engines for highly scalable database and analytics tasks. GPUs contain thousands of cores and have their own own memory, enabling high levels of computing power to be applied to offload processing in applications that lend themselves well to parallel execution.

“We’re one-tenth of the hardware cost and one-twentieth of the power consumption” of conventional relational engines running on specialized hardware, said Amit Vij (pictured), co-founder and chief executive of Kinetica. “We have Fortune 50 customers that are accustomed to getting bills from Oracle or SAP that are north of $100 million, whereas we cost just a few million.”

Kinetica was founded in 2009 to do work for the U.S. military in applications like real-time drone tracking. The company caught the attention of former Oracle Corp. President Ray Lane, who spearheaded the first round of fundraising. “They have created a database platform that acts as a data warehouse with advanced analytics and visualization,” Lane said in a prepared statement.

Unlike Relational engines that adapt themselves to GPUs, Kinetica was written from the ground up for GPU hardware, Vij said. “We started with a completely different data model,” he said. Kinetica calls its product a converged platform that supports things like machine learning libraries, visualization and advanced analytics natively, thus cutting down on integration costs. “We’re not duct-taping five or 10 different technologies together as organizations have to do with open source. We have all these capabilities in a single platform,” Vij said of the benefits of convergence.

Logistics and geographic applications continue to be a strong suit, but Kinetica said it’s also finding a home in genomics research, fraud detection and general-purpose data warehousing. Many customers also use it to accelerate popular analytics visualization tools such as Tableau Software Inc.’s, Vij said.

The new round of funding brings Kinetica’s total backing to $63 million and “gives us the buffer to operate for several years,” Vij said. The company was profitable when it was just a military contractor prior to hiring a sales and marketing staff. “There is some balancing we need to do, but we’re on a good pace that we’re pretty close to break-even,” he said.

Image: Kinetica