He has written many sportscar buying guides, a few mountaineering guides and drives a car that's faster than he is.
Comment Conventional wisdom has it that VC-funded tech start-ups get acquired or, hopefully, go through an IPO to become successful stand-alone companies. However, such may not be the fate of the five main surviving all-flash array (AFA) vendors in the face of fast and furious mainstream supplier reaction to their success.
Step back a moment and remember the last big round of storage array start-up disruption from 3PAR, Compellent, EqualLogic, Isilon, LeftHand, Nexsan, Storwize and others.
Basically, the good ones got bought, leaving Nexsan to stumble on, failing to go through an IPO and eventually getting acquired by Imation. Now, with a change of management, things are moving on:
- Dell bought Compellent and EqualLogic
- EMC bought Isilon
- HP bought 3PAR and LeftHand
- IBM bought Storwize
However, the same pattern is not happening with all-flash arrays, where four mainstream vendors have devised their own flash-array tech, with three buying their way in and one new enterprise storage array supplier doing likewise, leaving the five remaining AFA startups looking somewhat exposed.
- Cisco bought Whiptail and its Invicta array, now in temporary hibernation
- Dell all-flashed its Storage Centre (Compellent) array
- EMC bought XtremIO and DSSD and has all-flash VMAX.VNX arrays
- HDS all-flashed its VSP and HUS arrays with a HAF module
- HP all-flashed its 3PAR array with the 7450
- IBM bought TMS and so gained its FlashSystem product line
- NetApp all-flashed its E-Series and FAS arrays and is building its own FlashRay tech
- WD bought Skyera so its HGST sub could enter the market
This leaves Kaminario, Nimbus Data, Pure Storage, SolidFire, Syneto, and Violin Memory as a distinct group of FA suppliers all trying for glory by replacing disk arrays with their flash systems.
Syneto has a European heartland and can do well there. That leaves five small suppliers facing seven much larger mainstream suppliers, plus HGST.
There is now no mainstream storage array supplier lacking an all-flash array, meaning they have no need to buy one. This makes acquisition unlikely as a viable exit strategy for the stand-alone suppliers. One, Violin, has IPOed already and the other four have to IPO successfully so their backers can get their cash out.
What are their chances?
Kaminario, which claims a total of $143m, $68m of which was raised last year – is expanding its sales channel. The participants in the last funding round must have been convinced enough with its growth situation and prospects to have coughed up $68m.
Nimbus Data, with no VC funding, has been quiet recently; the last press release was eight months ago.
Pure Storage has massive funding, to the tune of $470m, and is recruiting EMC sales hot shots aggressively.
SolidFire, with $150m funding, has just announced a product update, unbundled its software for hyperscalers and announced a virtually unlimited flash wear guarantee. It says its revenues are growing 50 per cent quartet-on-quarter.
Violin Memory, already public, has refreshed management and a refreshed product range. Its revenues are running at $18m–$25m a quarter. EMC says XtremIO has a $1.2bn run rate, meaning $300m a quarter. Herein lies the problem.
Can Kaminario, Nimbus Data, Pure Storage and SolidFire all IPO successfully?
Virtually all the main storage array suppliers are finding demand in their base for all-flash arrays. At least four of the stand-alone AFA suppliers say they are selling well too. So what happens when the great current performance disk array-to-flash array transition nears completion?
How long does this have to run? When it does tail off, will any of the independents then be able to withstand the channel strength of the mainstream suppliers and their ability to integrate their AFAs with their other offerings?
Add in the parallel disk array replacement/upgrade competition from hybrid array start-ups Nimble, Tegile and Tintri – who are all growing strongly – and the looming entry of another US storage supplier into the AFA market and the conclusion seems inescapable.
There are too many AFA suppliers. Unless the independents manage to gain and sustain a technology advantage over the mainstream suppliers which can keep their installed base buying and customer count growing, they will surely see their revenue growth rates tailing off as their business enters maturity far sooner than it might have done ten years ago.
Because mainstream array suppliers have reacted to AFA start-ups so fast and so well, the future for the remaining independents looks, well, less bright.
Some would say they'd best hope to IPO quickly, say before 2017/2018, or else face their brightly growing stars sputtering out and becoming dim reminders of what might have been, with customer bases devoured by mainstream array supplier suns. (Don't say Coraid anyone...) ®