Does this mean the market is sour on upstart flash vendors? Or is Wall Street's response to Violin Memory just about Violin, as Freud's cigar is sometimes just a cigar?
Violin was the market leader in Gartner's 2012 flash array survey, but we should avoid reading too much into that single fact. Violin's total sales last year were $73.8 million, and the company managed to lose $109 million during that period. Sales are increasing, with the last six months showing $51.3 million in sales, but still $59 million in losses.
Several newer startups, including Nutanix and Nimble, are selling kit at a faster rate than Violin. Nimble claims a run rate over $100 million a year. Even direct competitor Pure Storage, which sold its first few units in 2012 when the survey was taken, is fast approaching Violin in size. Pure also managed to raise $150 million in venture capital in August, versus the $162 million Violin raised in its IPO.
All these vendors use standard SSDs, rather than develop several custom ASICs as Violin does, so they can concentrate on software development. Software development is both faster and less expensive than coding ASICs.
[As the SSD market heats up, Howard Marks makes his predictions for who'll come out on top in "SSD Vendors: Which Will Win?"]
While Enterprise Strategy Group's Steve Duplessie was right when he tweeted that the IPO didn't flop because Wall Street hates Violin CEO Don Basile, there is an impression that Basile, and the Violin board, are a bit fonder of Basile than Violin's performance would seem to warrant.
Basile's total compensation was reportedly about $18 million last year; while that may be reasonable for running HP or IBM, anything over 20% of gross revenues seems excessive to me.
In addition, Basile received a grant of 5 million shares of Violin stock, worth over $35 million even at the stock's low point Monday, for creating a liquidity event (taking the company public) within 60 days of July 31.
I, for one, am experiencing just a touch of schadenfreude in the losses Violin's initial public investors took as the stock fell. These aren't grandmothers and orphans being taken advantage of by Jeff Skilling or Bernie Madoff--these are Wall Street insiders who, because of their wealth and position, have the influence to get shares at the initial offering price.
Most of the time, their investment banker friends price the offering so they can buy at $9 and dump the stock to the greater fools placing buy orders on eTrade for $15 or so as the stock sky rockets on day one. That didn't happen this time around.
Violin faces several significant challenges if it is to cut its losses before the $162 million it raised in the IPO runs out in about 18 months. The biggest is its lack of software-related functionality. Violin arrays were designed strictly for performance, as its recent 2 million IOP announcement demonstrated. But, like a top fuel dragster, it sacrificed everything else to go fast.
Even basic data management features such as snapshots are provided in a Violin array through a custom version of Symantec's venerable Storage Foundation running on a pair of Xeon processor cards. The competition--from Pure and Solidfire to Cisco's Whiptail and EMC's ExtremIO--can deliver more performance than 98% of the market needs while also supporting VAAI and deduplicating data. At the same time, PCIe flash cards, which Violin does now make, are stealing much of the OLAP and HPC markets Violin was initially successful in.
Now that EMC, Cisco (via Whiptail), HP and Dell have all flash products good enough for 95% of buyers, Violin will have to make some really sweet music to stay in tune with its shareholders.Howard Marks is founder and chief scientist at Deepstorage LLC, a storage consultancy and independent test lab based in Santa Fe, N.M. and concentrating on storage and data center networking. In more than 25 years of consulting, Marks has designed and implemented storage ... View Full Bio