Additional Series A funding was provided by Robert Bosch Venture Capital, Yokagawa Electric, and Darling Ventures. The company also converted an additional $2.5 million of initial seed capital from March Capital and The Hive, a big data and artificial intelligence studio that helped create FogHorn.
FogHorn’s fog computing software stack works on the edge of the network and can run on an Intel or ARM-based PC device. This allows an enterprise to apply analytics and machine learning on the edge and not rely on the cloud. “We can do it in real-time as it’s being captured,” says David King, CEO of Foghorn.
King notes that IoT devices are producing a lot of data and that the economics of sending all that data to the cloud don’t always make sense. Sending data to the cloud also requires a lot of bandwidth, and there is latency depending upon the type of connection used.
By using fog computing, FogHorn can take that data at the edge and apply analytics and machine learning there. “It’s the difference between using a small compute device versus a big data center,” King says.
King adds that the company isn’t advocating processing all data at the edge instead of in the cloud, but that for some applications fog computing makes more sense. “Because of the real-time nature of this, it makes sense for time-sensitive data,” he says.
Lowering Cloud Costs
Predix, which operates as a platform-as-a-service (PaaS), has attracted more than 11,000 developers so far. GE expects Predix to generate around $15 billion in revenue by 2020.