Pure Storage developing converged systems

Interesting strategy at this stage of the game

Diversification before an IPO and before (presumably) profitability seems to dilute the laser focus they have enjoyed. Spending other people's money to go after lumbering, legacy competitors is one thing. Getting excited after beating up an old person, and then deciding it is time to jump in the ring with Nutanix, etc. is...an interesting strategy.

Are they going to hire an overlay sales force? New marketing? Pull people out of the field and re-train them? Selling storage boxes is not the same as selling hyper-converged solutions.

Storage people don't buy them. Server people do. It is a completely different game.

This reminds me of Violin entering the PCIe card business...and Isilon announcing that they were getting into the "fiber channel block" game. Both were looking to be acquired. Fusion IO announces that they are building shared storage arrays...example after example for why this should be worrisome.

For all the talk about building something enduring, it appears as though "The Vegas Exit" is what this is all about.

Pure isn't pure for much longer, I guess.

One Trick Pony No More?

Good luck with the pivot, they will need it.

Interesting Times

Wild success in one area may be giving them delusions of grandeur. This will add a list of very strong competitors who have a head start in this space. It will also spread Pure's R & D dollars more thin. Rumor is they're burning through their VC cash at a rapid clip and this would only accelerate that, further delaying an IPO. To be honest, their storage tech isn't much better than any other SSD vendor, so I'm not sure what competitive advantage they'd expect to gain in a converged architecture. Interesting times.

Hmm..is this endorsing more on the hyper convergence? Boy..good for Nutanix!

Either this is a stupid move or a brilliant move by Pure (given they are still a startup)!

hyperconvergence

not sure if hyper-convergence is enterprise ready yet.

the vision of a single global flash enabled optimized(dedup/compressable) namespace is not yet a reality with any commodity flash vendor .all optimization is done on as single node (which becomes the bottleneck for performance). If this technology " in its current form" is ported to hyperconvergence ,it will never scale to midrange or enterprise .