Cloud gateways largely exist because most cloud computing services are built around object-based storage systems to enable them to scale dynamically. However, most existing applications are built on top of traditional file systems. The gateway solves the problem by translating the file into a format that is compatible with the object-based storage system used by a cloud service provider.

This week cloud gateways are getting a lot more attention following the acquisition of TwinStrata by EMC, which was followed one day later by an announcement from Panzura that it will give away a free virtual cloud gateways along with 2TB of free storage on the Google Cloud Platform.

Panzura CEO Randy Chou said the cost of storing data in a public cloud is dropping to zero. The only time cloud service providers make any real money is when customers try to access that data across a network. The Panzura effort is designed to encourage more customers to store data in the cloud, under the assumption that Panzura and its partners can generate more revenue from customers trying to access that data across a wide area network.

Panzura still plans to charge for controllers that customers use to access multiple instances of a cloud. But if a customer only needs to store data to one remote instance of a cloud, then it doesn’t make much sense to charge for the gateway anymore, Chou said. In effect, there is a difference between a gateway and full-blown storage controller that could be used to manage storage across multiple cloud instances running on, for example, different zones on Amazon Web Services (AWS).

Like Panzura, TwinStrata, along with CTERA Networks, Nirvanix, Cloudian, Riverbed Technology and, following its acquisition of StorSimple last year, Microsoft, offer a cloud gateway.

EMC hasn’t divulged its plans for TwinStrata, but EMC already has an Atmos line of object-based storage systems that serves as a natural complement to the TwinStrata gateways.

IDC estimates that public cloud storage services will be a $12.2 billion market by 2016. How exactly that revenue is going to be generated is, of course, an educated guess. In a world where both the gateways to cloud and the actual storage are free, customers have to access a lot of data to get to those kinds of numbers.

The good news is that storage controllers are getting more sophisticated. Zadara, for example, has a controller that works across multiple public clouds and NetApp is not too far behind.

With 44ZB expected to be flowing through the Internet by 2020 there’s clearly money to be made in data storage both in and out of the cloud. The challenge going forward is figuring out not only what percentage of that data is actually going to land where, but just as importantly, how frequently that data needs to be accessed.