August 16, 2010

In Storage & Networking, Big Numbers In Dollars and Data

Today, at least for those of us who watch the enterprise space closely, the big news is that Dell Computer has offered to acquire Fremont-based 3Par for $1.15 billion, a premium of more than 80 percent over the company's stock price. In a world where much of the tech news is dominated by small companies taking money from angels, it's interesting to see the gulf between what it takes to grow a successful hardware company and the more ephemeral Web-based or application based companies that play significant roles on practically everyone's smartphone. And while I haven't talked about it too much on the blog, trying to keep a black and white separation between my day job for much of the last decade and my more hobby-oriented interests here, I've lived it, participating in one venture backed storage startup for more than 8 years, from 2001 to 2009, seeing companies raise, rise, fall and fail. In storage, the big winners, with few exceptions, can raise hundreds of millions of dollars before reaching break-even, and may be worth billions on the other side. Others may never find traction at all. 3Par, which took on tech titans like EMC and IBM, proved to have a winning formula.

There are three major truisms in technology. The first, and most well known, is that of Moore's Law, which while it has slowed in recent years, dictates that CPU processing speed increases at a regular clip while reducing in price. The second is that data storage capacities and densities are doubling at practically the same rate. Just look at the gigabytes or terabytes on your desktop or laptop hard drive and compare that with 5 or ten years ago. And the third is the speed of the network, both wired and wireless, increases - from the Kbps-rated modems of yesteryear to the fast-flowing networks of today, including 10 gigabit Ethernet on the client side and high speed Fibre Channel on the back end of many data centers.

These three advances mean simply this - more data can be created, shared, transmitted and stored more quickly than ever. Entire industries have been spawned around managing the data flow and storage, enabling branch office access to centralized data, deduplication and compression, load balancing and virtualizing the resulting complexity. If you watch consumer companies, such as Twitter, Facebook and Google, you probably see each of those companies creating new standards for global file systems and redundancy. You see them eschewing traditional storage companies and building their own devices in an effort to keep costs down as usage spirals upward. The trends are both amazing and incredible.

Back in January 2001, as Web 1.0 was crashing, I left a Web services company (eventually sold to Oracle) and joined a small company called Synaxia Networks, which later launched publicly to the world in March as BlueArc. At the time, comparable network attached storage devices from EMC and NetApp were capable of scaling to a then-massive 7 terabytes, and performance was not a metric either of them dominated. Our approach was simple - by converting aspects of the file system from software to hardware, we could dramatically accelerate storage. We scaled not to 7 terabytes, but to 225. We promised five nines (99.999% uptime) of reliability, and performance that was ten times the competition. And if we were less than that, everybody knows that two to five times the speed of the incumbent is still pretty darn good.

As we debuted, the press attention at the time was incredible - as our launch, backed by $30+ million in funding and our CEO being a former top guy at Compaq computer, gained massive attention. We had headlines in the Wall Street Journal and New York Times. George Gilder proclaimed that our product "imperiled" all software based storage devices, and after a successful debut at PC Forum, one reporter at said it was like offering crack to CIOs. Pretty heady stuff, and not unlike other dramatic booms seen from companies that captured the tech press's attention, including the currently hot Twitter and Foursquare, to those less successful, like Handspring and Transmeta.

But building a storage company takes a lot of real money. BlueArc, which raised another $20 million just last month, has raised $200+ million over its lifespan. 3Par, purchased today by Dell, similarly raised $100 million in 2001 (as we were raising $72 million) and others raised similar amounts. Cereva Networks, whose assets were later purchased by EMC, had raised $137 million and laid off 140 employees back in 2002 after not getting off the ground. Zambeel closed in 2003, having raised $66 million, but selling only a single system. Panasas raised $25 million in 2008, one of multiple rounds for the firm. Maxiscale raised $12 million before coming out of stealth. Pillar Data, funded largely by Oracle's Larry Ellison, is expected to have raised between $300 and $400 million alone. So when I hear tech reporters hem and haw about Web startups raising $10 or $20 million, it doesn't make me blink, considering the world of big dollars I've operated in for a decade.

So why the big dollars? Why are venture capitalists so willing to put such big bets into spinning disk and faster networks? Because when things go well, the customer benefits are very real, and the returns could be even better. Customers will pay top dollar to reduce the amount of time it takes to build special effects or bring pharmaceuticals to market. Fast network storage devices are key in mapping out the earth's terrain from satellites, and combing its ocean floor for potential oil deposits. Fast network storage is being used to collect mountains of data by the government, to simulate nuclear weapons' testing, and build next generation vehicles. And those companies that won't compromise on the speed of execution will buy from new storage startups not named IBM, EMC and HP.

That's why Isilon, a competitor to BlueArc during my time there, is worth more than $1.1 billion today, even after its own public struggles. 3Par earned its way to the discussion and is now cresting above $1 billion. Ocarina Networks, a client of Paladin, was purchased by Dell last month, for an undisclosed sum. Ocarina's competitor, Data Domain, was caught in a bidding war between EMC and NetApp, eventually going to EMC for more than $2 billion last year - simply with the promise of reducing storage capacity!

Today, some of the biggest debates in the Silicon Valley are around angels versus venture capitalists, and whether a $500k round can tip you from one side to another. Some of the best known Web startups today are begging for a $25 million acquisition by Google, or so it seems. FriendFeed, one of the biggest acquisitions by Facebook, was rumored to be "only" $50 million. But on the other side of the datacenter, it is an entirely new ballgame, where hundreds of millions of dollars go in one side, and you could get billions out the other end, or you could get nothing. Companies like 3Par, BlueArc, Isilon, DataDirect Networks, Panasas and others have put pressure on EMC, NetApp and IBM to innovate, and expand their product portfolios. Companies like Data Domain, Ocarina Networks and Permabit are working to optimize storage throughout the datacenter. Emulex, Qlogic, Brocade and Cisco are working on faster networks, cards, adapters and protocols to make sure data can go between client and server and back again at rates previously impossible, and everybody is betting on standards they hope will put them in the best spot.

So congratulations to 3Par for their fantastic exit and sale to Dell. Congratulations to Isilon for fighting a tough battle and living the life of a public company, worth $1 billion and up. It's fun to see companies and people I once saw as competitors, partners and allies, who I rubbed shoulders with at trade shows, and with whom I traded taunts on Twitter, taking things to the next level. There is no doubt in my mind that others will be good stories, and some will go the other way with spectacular flameouts, equally incredible to watch, but for much different reasons. It's a very different ballgame over here.

Disclosures: As a former BlueArc employee and investor, I own private equity stock in the company. In addition, Emulex is a current client of Paladin Advisors Group. Prior to their sale to Dell, Ocarina Networks was also a client of Paladin Advisors Group. Maxiscale was also a Paladin Advisors Group client in 2010. At times, I may seek to do business with or engage with many companies in this list, or their competitors.