Violin Memory CEO: Fusion-io IPO Just the Tip of Flash Iceberg

According to Violin Memory CEO Donald Basile, Fusion-io’s proposed IPO is just the start of what should be a very busy era for the flash space. And he should know: prior to taking the reins at Violin, Basile served as chairman and CEO at Fusion-io. Granted, Basile has a vested interest in just such an outcome, but he backs up his confidence with some sound insights. During a call this afternoon, Basile discussed all things flash, including the prospects for Fusion-io getting what it wants from its IPO, and how Larry Ellison helped assure the presence of flash in enterprise storage.

Regarding Fusion-io, Basile said that if investors buy into the company’s proposed IPO range of 10 to 20 times trailing revenue, that would be a “a very bullish sign” from investors that the enterprise flash market is about to blow up. Considering how small a company Fusion-io still is, it would be an even bigger coup, he added. But it’s certainly not an impossibility. As Basile noted, Fusion-io’s OEM deals with IBM, HP and Dell (deals that took shape under Basile’s watch) are very valuable, and the improvements in latency, density and energy efficiency that flash memory and SSDs can deliver are well known by this point. There’s a reason why Facebook has spent enough money over the past six months that Fusion-io noted its recent revenue reliance on the social-networking giant numerous times among the risk factors in its S-1 filing.

However, if Fusion-io made filed for its IPO with the intention of securing a buyer, as my colleague Stacey Higginobotham suggested, Basile thinks it could spur an acquisition trend similar to what we’ve seen in the storage industry lately. Especially if a server maker buys Fusion-io, its existing and potential OEM partners would likely start looking elsewhere to avoid doing business with a competitor.

Fusion-io and Violin Memory play in rather distinct spaces, though, although Basile says they do cross paths occasionally. Fusion-io sells SSD cache components to server makers that want to boost performance, whereas Violin Memory mainly sells primary storage arrays loaded with flash memory and/or even higher-performance DRAM. As for Violin, Basile said its world got a lot more interesting in December when Oracle CEO Larry Ellison very publicly validated the flash array storage model with the introduction of Oracle’s SPARC Supercluster, which packs 13TB of memory and 256TB of flash storage alongside more than a petabyte of hard disk storage and 108 16-core SPARC processors. Violin doesn’t do whole systems like Oracle does, but it is the only other vendor selling flash-based storage arrays, and it does have a partnership with HP aimed at rivaling Oracle’s dominance in the high-performance OLTP world. With Oracle’s Exadata business (under which the SPARC Supercluster falls) expected to contribute billions per year to Oracle’s bottom line, Basile likes Violin’s chances at striking it big.

Basile defines big for Violin in the billions, citing analyst predictions that have SSD and memory array spending potentially rising to $11 billion over the next couple years from its current sub-billion-dollar level. At this point, with only Oracle and Violin really in the market, Violin stands to take a big chunk of that cash. Already, Basile noted, his company has hundreds of active client accounts. Plus, Violin’s technology, and flash in general, isn’t relegated just to the storage layer. Juniper Networks, for example, is a strategic investor in Violin, seeing the potential for flash memory to help eliminate performance bottlenecks in the network, something Juniper is targeting with its new QFabric architecture. Juniper isn’t alone in seeing Violin’s promise, either; the company just closed a $35 million funding round.

Even with all the interest in flash, however, Basile doesn’t seem to ready to follow in Fusion-io’s footsteps and take Violin public, although he didn’t rule out the possibility of getting acquired. Although he thinks the company works well as an independent supplier of flash arrays, the recent acquisitions of Netezza, Isilon, 3PAR and other storage vendors hasn’t missed his radar. If large IT vendors decide they want something, he quipped, they generally can convince shareholders to agree to a sale. If flash really is set to take off into the stratosphere, it seems only logical that someone is seriously considering snatching up Violin and keeping its technology to itself.