CIA’s Investment Partner Invests in GPU Analytics Startup MapD

In-Q-Tel (IQT), the venture capital arm of the CIA and the broader intelligence community, has infused an unknown amount of money into MapD, a startup that offers GPU-accelerated analytics software. IQT joins NVIDIA, Vanedge Capital, Verizon Ventures, and GV (formerly known as Google Ventures) as co-investors in the company.

IQT Managing Partner George Hoyem explains the rationale for their investment thusly:

“MapD is one of the newly emerging software companies leveraging advanced GPU hardware to deliver very impressive performance improvements in data visualization and data exploration,” said George Hoyem, Managing Partner, Investments, IQT. “The ability to interact with and visualize billions of data elements in real-time is a transformative capability for our national security partners.”

Back in March, MapD introduced its first product into the market, a unique GPU-powered database and visual analytics platform. The software grew out of a graduate project conducted by Todd Mostak at Harvard, who was attempting to study the role of Twitter on the Arab Spring. Frustrated by the inability of conventional database technology to deal with such a large and unstructured dataset, Mostak joined MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), where he developed a prototype of a new parallel database and visualization tool that could handle such tasks. In 2013, Mostak, along with co-founder Thomas Graham, founded MapD, where they commercialized the technology.

The software employs GPUs as a parallel engine to accelerate queries on SQL-compliant databases, leveraging the massive parallelism of the graphics processor and the high bandwidth of the graphics memory to execute hundreds of query threads simultaneously. Hot data is cached in the GPU's RAM. The graphic processors are also utilized to drive visualization for displaying the results of queries interactively. The result, they say, is a system that lets you interactively explore multi-billion row datasets.

In conjunction with the introduction of MapD’s first offering, the company also announced they had attracted $10 million in Series A funding from the aforementioned investors.  Even at that early date, the company claimed they had chalked up some early adopters. One of those is Verizon, which is using it to search for certain patterns and anomalies across a streaming log. Simulmedia, a provider of targeted television advertising, employs MapD to do predictive analytics across advertising datasets. Finally npm, which manages the open source Node.js Javascript engine, is using MapD to query and visualize billions of log records per month.

Such early traction will be critical for MapD, given the growing competition in this area. Over the last several years, a number of startups have built high performance analytics platforms powered by graphics processors.  In particular, companies such as SQream Technologies, BlazingDB, and Kinetica (formerly GPUdb), are all offering their own renditions of GPU-powered databases.

The investment by IQT bodes certainly well for MapD, since it implies there is serious interest by the generously funded US intelligence community. These three-letter agencies maintain some of the largest datasets in the world and have entire supercomputing datacenters devoted to their analysis. In this case, the investment looks to be targeting additional development for IQT’s client base.  According to MapD:

“IQT approach is designed to accelerate product development and add mission-critical capabilities with the sole purpose of delivering cutting-edge technologies quickly and efficiently to their client base. We are currently in what IQT identifies as the “work program” phase where we adapt, extend and deliver our software to an IQT client.”

To read more about the news, go to the MapD blog that details IQT’s engagement.

Currently unrated


There are currently no comments

New Comment


required (not published)