Nimble flashes the all-flash array as ‘intense’ consolidation period approaches
Hey, Tegile and Tintri. Are you ready for this?
The scaling challenge
"The next leg of the journey, which we would speculate will be in the first half of calendar 2016, will be a true AFA, where the data will be persistent in flash and no disks will reside in the system," he added.
He senses that Nimble will differentiate its AFA, "some tricks up its sleeve, which management is not yet comfortable talking about".
Ader also says Nimble is indirectly active in the converged/hyper-converged market space, and “is seeing strong traction partnering with Cisco for its SmartStack converged solution ... Ultimately, Nimble believes that a maximum of 25-30 per cent of the storage market will move to converged/hyperconverged architectures".
The main challenge is around scaling, Ader writes, “which becomes very expensive as the need to replicate data across multiple nodes requires overp-rovisioning the number of nodes".
In addition, larger data centres typically do not want or need to scale equally between compute and capacity, which a hyperconverged system requires (though products such as Nutanix offer the ability to deliver compute-heavy or storage-heavy nodes).
According to Nimble, hyperconverged works best in more uniform environments where simplicity matters most (eg, remote offices, smaller data centres). In higher scale, less predictable environments, hyper-converged is not a preferred solution, according to Nimble.
He thinks Nimble will do well: “We continue to see Nimble as one of the winners in the next-generation storage space.”
While “Nimble could be an attractive acquisition target given our belief that a period of intense consolidation is likely to take place in the storage industry over the next couple of years as incumbents have no choice but to acquire their way out of the problem".®