Cloud infrastructures are the back-end to the cloud. They are what the providers of the customer facing cloud service will eventually host their applications or services on, store their data on and they are what local applications like backup and archive software will augment as an additional tier of storage.
There are basically two types of infrastructure providers today. Companies that have set out from day one to deliver cloud compute and/or storage. These are companies like Nirvanix, Amazon, Google and Microsoft. This group of companies essentially created their own storage and/or compute platforms to deliver cloud services, in many cases creating their own storage software solutions.
The second group are those that really provide a cloud enabled service but have decided to host the compute and storage themselves. As I said in my prior blog I see this second group moving to get out of the infrastructure business and leveraging the infrastructures of the first group or partnering with the next wave of infrastructure providers.
The next wave of infrastructure providers will more than likely be traditional service providers, companies that understand facilities and logistics, not software developers. As a result I see them using more off the shelf solutions like EMC's Atmos or Cleversafe. They may also take software solutions like those from Bycast or Mezeo and then use focused storage hardware like Permabit's or HP's Cloud Storage offerings to deliver a complete solution.
This next wave of providers could potentially be a huge market of new customers for both storage hardware and software manufacturers. As I state in our video on Cloud Storage Infrastructures, many suppliers will try to take advantage of this market by repackaging "cheap disk". Infrastructure providers need to be wary of this approach and make sure that the storage component of their infrastructure will be reliable, scalable and affordable throughout its use.