everythingpossible - Fotolia
Will the flash storage market see another big acquisition?
HPE's $1.2 billion deal for Nimble Storage has stirred speculation on potential future acquisitions in the all-flash array market, such as Kaminario and Pure Storage.
Hewlett Packard Enterprise's recent $1.2 billion deal for all-flash and hybrid array vendor Nimble Storage caught many market observers by surprise and ratcheted up speculation about potential future acquisitions.
But flash storage market analysts said the chances of another big acquisition in the all-flash array (AFA) space are a long shot, and they don't expect the remaining early flash array vendors to cash in soon.
The initial wave of startups that built storage arrays exclusively for NAND flash has dwindled to two significant players: Kaminario and Pure Storage. Pure Storage is an unlikely acquisition candidate because of its market cap. Pure, which went public in October 2015, ranked fourth in revenue in the AFA market in 2016 behind Dell EMC, NetApp and Hewlett Packard Enterprise (HPE), according to IDC statistics for external storage arrays that have a unique all-flash SKU and use only flash media.
Kaminario, a private company, does not publicly disclose its revenue, but IDC has access to it. Eric Burgener, an IDC storage research director, said Kaminario was sixth in the all-flash storage market for 2016, with a substantial lead over seventh-place Hitachi Data Systems (HDS), but far behind No. 5 IBM.
"They're not a small company anymore," Burgener said of Kaminario. "I can't imagine who could afford them that wants to get into this space and isn't already there. That doesn't mean they're not trying to be acquired."
Kaminario on the block?
An industry source who asked not to be identified claimed Kaminario's venture capitalists have approached server vendors. However, Kaminario CEO Dani Golan said the company is not actively seeking a buyer and is not receiving pressure from its investors to do so.
Eric Burgenerstorage research director, IDC
"The only thing that we can do as a company is to execute and not worry about anything else," Golan said. "We have some of the best financial institutes behind us, and they all say, 'Go do the right thing.' And that's really what I'm focusing on and nothing else."
Kaminario secured $75 million in financing in January, lifting the company's total to $218 million for five funding rounds since 2008. Golan said the latest round was heavily oversubscribed, and investors were excited to see Kaminario focus on infrastructure for the growing market of software as a service and public cloud.
"We chose to grow responsibly -- not grow like crazy and burn money," Golan said. "It doesn't put us in a corner or rush to sell the company. We invested in the technology long term."
But, Tim Stammers, a senior analyst at 451 Research, said Kaminario may have missed its chance. "Kaminario took too long to come to market; the market is now mature, and the [large storage vendors] have locked it up again," he said.
Flash storage market changed
Stammers pointed out large storage vendors have already picked up most of Kaminario's early flash storage market rivals. In 2012, EMC bought XtremIO and IBM bought Texas Memory Systems. Western Digital's HGST subsidiary purchased Skyera in 2014, and NetApp scooped up SolidFire in late 2015.
Cisco, Lenovo and Huawei are the major server vendors who observers have said may be interested in entering the flash storage market. But Cisco already spent $415 million on AFA vendor Whiptail in a 2013 deal that went bust when the product was pulled after generating little revenue. Lenovo has been getting more active in storage but -- like Cisco -- tends to focus on partnerships. Huawei, a Chinese company with a strong presence in Asia, already sells AFAs but has struggled to break into the U.S. market.
"Any vendor that's going to take on the top five vendors in what's already a pretty mature market is going to have to be satisfied with being a niche play," Burgener said. "I don't see Huawei buying Kaminario and turning that into a top-five market share vendor."
Marc Staimer, president of Dragon Slayer Consulting, said he could not envision any server vendor buying a flash storage array vendor, whether it's Kaminario or Pure Storage, or all-flash or hybrid vendors Tegile Systems or Tintri. Investors have poured hundreds of millions of funding into each one of those vendors, and none of them are profitable.
"They've all had too much money put into them, so the investors are going to want more than the market will bear," Staimer said.
Growing flash storage market
Staimer said the all-flash array market is growing, but the overall external storage array market is declining. "So if I'm a server vendor, do I want to buy into a shrinking market? Really what I want is a software-defined storage play that fits with my servers," he said.
Kaminario's Golan claims his company's K2 product is "completely software defined" and can run on any commodity x86 servers, although it currently bundles and tunes its product only for Supermicro hardware.
"They wouldn't be acquired as a software player," Staimer said. "Just because they run on commodity hardware, it doesn't mean they can be bought and integrated into somebody else's hardware. That's not a trivial process."
Staimer said Hitachi Data Systems might be a more likely candidate to acquire Kaminario because HDS has fared the worst among the major storage vendors in the AFA market. But, he said, HDS may not have the cash, the stock or the inclination to pursue a deal at this point.
George Crump, founder and president of Storage Switzerland LLC, said more interesting potential targets for HDS and possibly even IBM might be Pure Storage, Tegile or Tintri over a high-end AFA vendor like Kaminario. "I think Kaminario's a great product. I just don't see a good match there with anybody," he said.
New rules for storage startups, acquisitions
Crump said it's difficult to predict what will happen in the storage market today, unlike the past when one could foresee HPE acquiring 3PAR or EMC buying Data Domain to fill gaps in their portfolios. After the world's largest storage company, EMC, got acquired by Dell for more than $60 billion last year, no deal can be dismissed.
Market conditions have also changed the way startups are run and perceived. The days are gone when a startup could burn through large sums of funding to grow its business to attract suitors without watching the bottom line.
"Today, you need to build a company with the intent of being profitable, and if at some point during that journey you have an opportunity to go public, go for it," Crump said.
Even storage companies that go public often get acquired. Data Domain, 3PAR, Isilon, Compellent, Fusion-io and Nimble Storage all went from startups to public companies before larger public companies gobbled them up. Others, including EqualLogic, were bought while preparing an initial public offering.
"If you ask any startup, 'What's your exit strategy?' they will say, 'Oh, we're not aiming to be acquired. We're aiming to build a good strong business,'" 451 Research's Stammers said. "But name the startup that has survived without being bought. The last big independent storage maker that came out of nowhere? NetApp. Slightly before that, EMC. We have Pure and Nimble, but Nimble wasn't making a profit. Pure is not making a profit. It's really hard to make it in this market."
AFA pioneer Violin Memory this week completed its restructuring and sale to Quantum Partners, a private investment fund run by Soros Fund Management. Violin went into Chapter 11 bankruptcy in December 2016 after failing to find a buyer when the company's CEO put it up for sale in 2015. Quantum put up less than $50 million for Violin at auction.
Stammers said the new wave of ultra-high-speed, analytics-focused NVM Express-based AFAs might be more likely to produce acquisition targets down the road than the remaining early players in the flash storage market. Vendors include Apeiron, E8 Storage, Excelero and Mangstor. Stammers noted that Dell EMC killed its DSSD product, claiming a lack of demand.
"But that might not be true. It may be because DSSD was too expensive. It had a lot of custom hardware in it," Stammers said. "If analytics drives big demand, we might see the OEMs going and buying a second wave of products. It's also quite possible that EMC is right. There is no market here at the moment. It could be a bust."