The storage-focused BD Event wrapped up last week in Boston after two and a half days of presentations and networking for the enterprise storage industry–executives meeting up with executives in a kind of industry-focused, back to old college days mixer. If you’re an entrepreneur with designs on the enterprise storage market, this was the place to be.
The event was bracketed by two provocative presentations: one delivered by Peter Bell of VC firm Highland Capital Partners to kick-off the event, and one by Peter Levine, senior vice president of Citrix, that concluded the formal presentation session the following day.
Bell’s perspective on the enterprise storage industry is always worth taking in. As a former EMC executive and later chief executive of Storage Networks, the highest-flying of the Internet V 1.0 storage service providers, he has an insider’s perspective on some pivotal evolutions–from monolithic to modular array storage, and from storage utility to storage cloud.
There were a set of mixed messages here. Yes, there are opportunities, but storage start-ups face a challenging fund-raising climate. Venture capital firms are placing increasing emphasis on capital efficiency and are looking to see product development and revenue milestones faster. The climate can be particularly difficult for start-ups whose products have a hardware component that require more capitalization than pure software plays. As a result, storage-focused entrepreneurs are increasingly seeking help from other funding sources including angel investors. The good news is that, overall, we are heading into an increasingly optimistic technology investment environment.
The concluding session was delivered by Levine, who has an insider’s view on virtual-desktop infrastructure (VDI)–currently one of IT’s hottest topics. Levine had an answer for audience members who might have been left wondering, after Peter Bell’s pitch, where the best opportunities are. Levine listed a set of requirements for centralized data center storage intended to support VDI environments. Currently he sees enterprises rolling out pilot VDI deployments of 2,000 or so desktops in organizations with the potential for 50,000 or more. He is concerned however that the high cost of tier-one storage arrays will become an inhibiting factor as data migrates from the desktop to the data center.
As he sees it, because the cost of a gigabyte of storage on the desktop is much lower than the same gigabyte on a tier-one data center array, IT administrators will have increasing difficulty justifying larger VDI deployments that move data from application user-land into the embrace of centralized IT. Making the cost of data center storage roughly equal on a dollars per gigabyte basis to that of storage on the desktop removes a barrier. What’s needed? Highly scalable, object or file-based storage that’s also significantly cheaper than today’s gold-standard, production disk arrays. Levine invited audience members to contact him after the session with ideas.
While VDI may well be the way many IT organizations eventually spell “private cloud.” VDI will challenge IT administrators in the early going to justify the initial expenditures. The gold-standard storage vendors will argue for a highly performance-oriented storage layer that has the ability to scale capacity quickly as virtual desktops are added by hundreds if not thousands of users in a single rollout cycle. The software that manages the VDI storage layer will also have to be integrated with other layers in the VDI support stack and support near instantaneous disaster recovery capabilities. Today that’s premium stuff, to say the least. (One member of the audience observed that a leading vendor of premium storage didn’t buy VMware on a mere whim.) The challenge and the opportunity for start-ups will be to develop storage infrastructure components that compete with the established vendors at a price for performance discount never before seen.