There's a lot of buzz surrounding the growing list of startups looking to break into the enterprise storage industry, so much so that questions are starting to be asked regarding the long-term prospects of top-tier storage vendors like EMC and NetApp.
The catalyst in all this is Flash storage. As costs come down and capacities go up, many of the newcomers are looking to wedge their way into enterprise settings with all-solid-state storage arrays that push IOPS into the stratosphere — where the clouds are.
Newcomers like Nimble Storage and Pure Storage are drawing not just customers toward their platforms, but venture capital as well. Besides faster throughput and increased capacity, their selling points include streamlined deployment, faster backup and recovery and improved analytics that gauge the health of both the storage array and the surrounding network. And most ominously, these platforms are starting to come in at less cost than the mechanical-drive solutions of the EMCs and IBMs of the world.
Startups also have the advantage of not having to protect legacy architectures. To their credit, EMC and IBM have recently acquired some innovative new Flash technology — IBM through its purchase of Texas Memory Systems and EMC's buyout of XtemeIO — and there is still time to assert their leverage with leading enterprise customers looking to increase their reliance on solid-state storage. But technology cannibalism is not as easy as it sounds, and the need to gain market share in all-Flash arrays will likely wage a constant battle against the desire to foster hybrid solutions that allow them keep a foot in both doors.
Indeed, while startups are gaining deployment wins with all-Flash platforms already in the channel, EMC is still in the demonstration phase of its XtremeIO-fueled Project X. The company brought it to VMworld last month, where attendees saw a 4 RU, dual-controller machine with dedicated Intel CPUs for each drive. EMC has been sure to note that with its Flash arrays, users will still benefit from what many now consider to be standard storage functions like thin provisioning, snapshots and the like. A formal release is expected sometime in 2013.
NetApp is also adding Flash to its portfolio, but appears at this point to be focused mainly on building high-speed cache to help mechanical drives handle critical data. The company has tapped Fusion-io to bring its server-side platform into the Flash Accel storage management software realm, providing an integrated management stack aimed at improving application performance some 90 percent. This will no doubt be a welcome development for many NetApp customers, but it falls short of the all-Flash arrays that are fueling such high interest in the startups. Also, keep in mind that technology partnerships usually last only when both partners are fully independent businesses, and EMC could probably do wonders with technology like Fusion-io's if it were to keep its acquisition streak alive.
Speculation like this can veer down many paths, however. It could be that all-Flash storage will prove to be a flash-in-the-pan, or that long-standing partnerships between enterprises and storage vendors will be enough to quell the startups' momentum before too long.
Then again, it could be that we are on the cusp of a technology disruption that will have broad implications in the storage market and beyond. It wouldn't be the first time that changing expectations led to changes in enterprise infrastructure.