NVIDIA : Powered by GPUs, Real-Time Data Startup MapD Lands $25M Funding Round

That was fast.

Just a few years ago, MapD had a radical idea: Use GPUs to turn data into insights, quick. The result was a jaw-dropping demo. And a pitch that won the then-tiny startup a $100,000 check - awarded on the spot - at our Early Stage Challenge in March 2014.

Almost three years to the day later, MapD has closed a $25 million series B funding round. Now a 30-person dynamo, the San Francisco-based startup plans to accelerate its growth as some of the world's largest companies put GPUs to work on their toughest business problems.

It takes only a moment for Todd Mostak - a Harvard-trained expert in Middle Eastern politics who founded MapD while doing a fellowship in AI at MIT - to reel off the opportunities. The data warehousing market is worth $30 billion. Business intelligence and visual analytics are another $20 billion. And the size of the emerging AI market looks 'humongous.'

'We play across all of these,' Mostak says.

MapD's GPU-powered software brings two radical new capabilities to market. MapD can use the massively parallel processing capabilities of GPUs to instantly query multi-billion datasets. That lets businesses answer queries that once took as long as hours in an instant. And because of the GPU's roots in visual computing, MapD can turn all these billions of pieces of data into rich, interactive visualizations.

None of this would be possible without GPUs. Mostak says.

'People have gotten accustomed to incremental and piecemeal gains. By contrast GPUs are delivering 50 percent year-over-year performance gains,' Mostak says. 'These aren't small steps, these are giant leaps, and we're riding that trend.'

To keep at the forefront, Mostak will use the money raised from its latest round of funding - led by New Enterprise Associates, with participation by NVIDIA, Vanedge Capital and Verizon - to expand MapD's sales and marketing team and add customer service representatives.

MapD's technology, meanwhile, is still dropping jaws. Verizon, for example, is using MapD's software - and GPUs - to get quick insights into the performance of the company's vast infrastructure.

And the idea of turning GPUs into internet-connected, data-crunching engines has gone from a radical notion to a capability any business can turn on in an instant. Amazon, Google, IBM and Microsoft, among others, have all launched cloud services that make it easy to put GPUs to work.

'GPUs are at the front of everyone's consciousness right now,' Mostak says. 'It's an idea whose time has come.'