Top Ten Private Companies: Summer 2003

It's hazy, hot, and humid stickier than a jam jar full of angry bees. Yes, we're talking about the Byte and Switch Top Ten Private Companies list, making a summer appearance and wooshing up a gentle breeze over the sweat-stained storage networking industry.

To recap: The Top Ten list is based exclusively on our evaluation of the market and the prospects for each company's long-term success. To be considered, companies must be privately held, have shipping products, and demonstrate the existence of revenue-producing customers. Bonus points are given to startups that do not have foosball or ping-pong tables on the premises; points are deducted from those with humorless, pinch-faced executives (see our previous list, Top Ten Private Companies: Spring 2003).

Here are the highlights of the summer's reloaded Top Ten list – iced down, oiled up, and sun-kissed for your beach-reading pleasure:

Before we proceed, we should remind readers that, unfortunately, there is room for only ten companies here. This isn't a Top 5,000 list, amigos. Keep in mind that just because your company missed the cut doesn't mean we don't like its chances. And by all means, throw your own picks and pans onto our message boards (but please clean up after yourself).

For a writeup on each company, click on its name below:Table 1: Top Ten Private Storage Networking Companies

— The Editors, Byte and Switch

Clambering back to the top of the heap is NAS vendor BlueArc Corp., which convinced its VCs to stuff a whoppin' $47 mil into its front pocket. (Fresh!) The money, says BlueArc, will last until it hits profitability (see BlueArc Wallows in $47M Haul).

The VCs were persuaded to give BlueArc the booster shot based on its fairly impressive customer traction, which started to accelerate in the second quarter. Helping on this front is that BlueArc can now offer lower entry price points with ATA disk options, meaning a customer doesn't have to bet the ranch to acquire one of its NAS systems.

And check this out: It's gotten into not one, but two Ivy League schools! Recent customer wins include:

Need more proof BlueArc has arrived? In the most backhanded of compliments, Network Appliance Inc. (Nasdaq: NTAP) is suing the company for allegedly infringing patents that NetApp bought from bankrupt Auspex Systems. (There's a neat trick, eh?) While this isn't a positive development, because it has the potential to distract BlueArc, it shows that NetApp clearly sees BlueArc as a competitive threat. Perhaps lawsuits are the sincerest form of flattery (see NetApp Zaps BlueArc With Lawsuit).

One concern about BlueArc is that its architecture, which accelerates NAS transactions in silicon, will hit a wall at a certain point – one huge box can only scale up so far. On the other hand, NetApp has been successfully riding its own software-based NAS architecture for 10 years and doesn't show signs of slowing down. Onward and upward, BlueArckers! Let's see how long you can stay in the top slot this time.

Coming on strong in second place is NuView Inc., whose metadata software provides a global namespace for NAS servers.

We continue to like NuView's proposition: Simple installation; not incredibly expensive; doesn't force customers to revamp their entire NAS infrastructure; smells nice. In a nutshell, it solves a customer problem and does so very neatly.

Another major plus for the startup is its tight alliance with Network Appliance Inc. (Nasdaq: NTAP), which resells NuView's StorageX software. In fact, we hear that it's pretty much just a matter of time before NetApp snaps up NuView – not a question of if, but when. Note that NuView founder and CEO Rahul Mehta has sold three companies prior to this one, to Computer Associates International Inc. (CA) (NYSE: CA), Hewlett-Packard Co. (NYSE: HPQ), and Veritas Software Corp. (Nasdaq: VRTS).In the nonce, NuView continues to sign up new customers and enhance its product. The latest release of its software, StorageX 4.0, adds the ability to move data among NAS devices automatically based on criteria like file size or type of data (see NuView Relocates NAS Data). But never mind that: NetApp thinks NuView offers simple, powerful software – and NetApp should know.

CommVault Systems Inc. charges up three places, based on its continuing execution as well as its vision for an integrated data backup-and-recovery platform (see CommVault Counts 495 Wins for 2002).

The latest feather in CommVault's cap is an OEM deal with Storage Technology Corp. (StorageTek) (NYSE: STK), which plans to offer a StorageTek-branded version of CommVault's quick recovery application. For CommVault, it represents a major new channel and a key partnership with one of the industry's biggest players (see CommVault Packs Up OEM Deals). Note, too, that CommVault counts Microsoft Corp. (Nasdaq: MSFT) among its investors.

Like the other companies on the Top 10, CommVault is tightly focused on a specific market sector; in this case, data backup and recovery, an area that has been growing fast and is increasingly important to businesses. Larger competitors – like Veritas Software Corp. (Nasdaq: VRTS), IBM Tivoli, and, soon, EMC Corp. (NYSE: EMC) via its acquisition of Legato Systems Inc. (Nasdaq: LGTO) – have much broader strategic goals. That leaves enough room for a specialist like CommVault to zero in on the large universe of midrange enterprises that simply want reliable, useful data-recovery software.

We hear rumblings now that CommVault is considering going public sometime in the near future, a notion it's been kicking around for well over a year (see CommVault Aims for IPO). If it does... well, it will come off our list. But that's one of the preferred methods of exiting the Top 10.DataCore Software Corp. inches up a notch, as it maintains a steadfast focus on delivering heterogeneous SAN management software and keeps on racking'n'stacking customer wins and partners (see NEC Soft OEMs DataCore, DataCore Nurses SANs, German Chemists Pick DataCore, and DataCore Boots Up Printer).

While several larger players (including erstwhile DataCore partner IBM Corp.) are futzing around with newfangled network-based virtualization appliances and switches, DataCore is finding a home for its Windows-based software, which now takes advantage of storage-centric features in Windows Server 2003 (see DataCore Boosts SAN Software).

The wild card is that DataCore is delivering software that runs on a PC – which is an extremely tough sell to, say, a large Unix shop. The knock against this approach is that it won't scale. But DataCore keeps on proving that what many customers want and need today is a centrally managed, single point of control for their storage. It will happily leave the high end of the virtualization space to the hardware guys; software is by and large a more attractive business model.

DataCore, like CommVault, is also supposedly pondering an IPO at some point, but we shall see where the market takes it. For now, it remains on our list because it's playing by its own rules and managing to win.

Making its debut at No. 5 is NSI Software, a 12-year-old geezer on this list of young'uns.Until recently, NSI was a relatively sleepy little software company nestled in the paradisaical splendor of Hoboken, N.J. In the past two years, however, it has parlayed its Double-Take data replication software for Windows into a high-growth engine. To date, the company says it's sold 16,000 server licenses operating at more than 5,000 customer sites. NSI has seen eight consecutive quarters of revenue growth, and CEO Don Beeler claims it's just starting to hit the curve in the hockey stick.

What's behind the surge? [Ed. note: We know for a fact it can't be the Hoboken nightlife.] For starters, NSI's software costs around $2,000 per server. By comparison, the starting list prices for EMC Corp.'s (NYSE: EMC) Symmetrix Replication Data Facility (SRDF) disk replication software are more than 10 times that. "What's affecting our market more than anything is the slow economy, and right now that's affecting us in a positive way because you can just buy what you need," Beeler says.

NSI also has distribution deals with Dell Computer Corp. (Nasdaq: DELL) and Hewlett-Packard Co. (NYSE: HPQ), and recently inked an agreement with SunGard (NYSE: SDS), a provider of offsite data hosting services, which plans to launch a service using NSI's software. The company counts around 50 resellers worldwide.

At a higher level, there are two factors propelling NSI: First, after 9/11, enterprises have never been more cognizant of the need to replicate their critical data; second, most servers today run Windows. As Emeril Lagasse would say: Bam! NSI is smack in the middle.

The 150-employee company is almost at the breakeven point ("We have profitable months," says Beeler) and appears well set on the funding front. In January 2003, NSI closed a $15 million second round, which included an investment by Dell, bringing the total raised to $65 million to date (see NSI Notches $15M).However, we should note that NSI will continue to face serious competitors like Veritas Software Corp. (Nasdaq: VRTS) on a daily basis, and it will soon be directly vying for accounts against EMC, which is buying Legato Systems Inc. (Nasdaq: LGTO), another of NSI's traditional rivals. Not incidentally, Legato is suing NSI for alleged patent infringement (see Legato Lobs Lawsuit at NSI). And there's always the danger that Microsoft Corp. (Nasdaq: MSFT), as it has a tendency to do, will simply suck up data replication features into the operating system at some point.

But it's come this far, and NSI looks to have all the right body parts and metabolism to keep on going.

Drifting down from its previous No. 1 spot, LeftHand Networks is downgraded because it appears that its true differentiator – providing virtualized pools of IP SAN storage – will soon be evaporating.

So far, LeftHand has garnered plenty of customer traction and managed to scoop up a nice basket of $20 million in second-round funding earlier this year (see LeftHand Snatches $20M). Nothing sinister about that. But as standards-based iSCSI technologies begin to proliferate in the marketplace, LeftHand's Ethernet storage system, which uses a proprietary Ethernet-based protocol, will begin to look less attractive. The market will move toward standards, regardless of LeftHand's insistence that its own Advanced Ethernet Block Storage (AEBS) protocol is superior to iSCSI.

LeftHand does have plans to adopt iSCSI in its offering in early 2004. But at that point, it will be staring down the barrels of the big guns of major storage systems companies, including Dell Computer Corp. (Nasdaq: DELL), EMC Corp. (NYSE: EMC), and Hewlett-Packard Co. (NYSE: HPQ). It will also have to compete for business against other startups pushing iSCSI-based arrays like EqualLogic Inc. and Intransa Inc.As an IP SAN pioneer with some early success, LeftHand deserves to stay on our Top 10. How successfully it can move to catch the rising iSCSI tide remains in question.

ManagedStorage International Inc. (MSI) hangs tight at No. 7, demonstrating that, with the right approach, a variation on the storage service provider (SSP) model can work just fine.

Tom Sweeney, MSI's president, CEO, and chairman, says the company racked up $6 million in managed storage service contracts in the second quarter, primarily for backup and recovery operations. Meanwhile, StorageNetworks Inc. (Nasdaq: STOR), one of the original SSPs, fell on its sword last week, announcing that it was dissolving in order to return some money to its shareholders (see StorageNetworks Succumbs). In fact, MSI has acquired about nine former StorageNetworks' managed services customers after the latter decided to exit that business in June (see StorageNetworks Ditches Services Biz).

"The managed storage service business is one that does work, but it requires an effort where you have multiple distribution channels, and you have to deal with both large enterprises and small customers," says Sweeney. MSI currently has around 70 employees and serves 650 customers; its revenues are split 50-50 between large enterprise customers, and service providers (which resell private-label versions of its services) and small and midsized businesses.

MSI has also been growing internationally, with data centers in Tokyo, London, and Paris. It recently signed a partnership with London-based Orthus Ltd., an information security services firm that will resell MSI's data protection services, and Sweeney promises more such partnerships later this year. Stay tuned.Creeping up a spot is GlassHouse Technologies Inc., a SAN consulting boutique that is ably exploiting the growing need for expert storage networking help.

GlassHouse flies well below the radar of giants like Electronic Data Systems Corp. (EDS) and IBM Global Services. While it doesn't generate hundreds of millions of dollars in revenue, GlassHouse has been carving out a respectable niche by serving clients as a trusted, strategic advisor. It doesn't sell hardware or software. Where the big boys sell bodies [ed. note: figuratively], GlassHouse sells brains [ed. note: again...]. (See Guardian Taps GlassHouse, GlassHouse Boosts Brosco Backup, Softek Opens Door to GlassHouse, GlassHouse Wins Hartz, and GlassHouse Wins in Wash. State.)

We also like the fact that GlassHouse opportunistically scooped up 330 support contracts of failed NAS vendor Auspex Systems for a token $280,000 – acquiring several hundred high-end customers for less than $1,000 apiece is a nice move (see GlassHouse Picks Auspex's Bones). As it continues to grow organically, opening new offices recently in New York and Minneapolis, GlassHouse (clearly) is still one to watch.

3PARdata Inc. keeps a spot on the list if only because it's the last viable large-scale, next-generation storage array startup left in the industry. (Cereva who? Yotta what?)

Of course, being the last-man-standing is no cause for celebration. But to win against the large incumbents, 3PAR has been forced to demonstrate that its system is faster, more scaleable, more flexible, and more affordable than storage from EMC, HDS, or IBM. This is an exceedingly tall order. And yet, 3PAR has managed to do that, steadily picking up customers – including, much to 3PAR's immense delight, a division of Hitachi Ltd. (NYSE: HIT; Paris: PHA) (see 3PAR Hits Up Hitachi and 3PAR Gets to Infinity).Another tool 3PAR now has in its belt is Thin Provisioning, a twist on the "capacity-on-demand" concept that lets customers set up a volume without buying the full amount of storage required. 3PAR expects this to lower the upfront cost of its storage array, thereby greasing the skids for customers to put in purchase orders. It's hard to convince those pesky bean counters to write a million-dollar check to a four-year-old storage startup, isn't it? (See 3PAR Spins Disk Trick.)

It's not certain, however, how successful Thin Provisioning will prove to be for 3PAR [ed. note: Hey! It's worked for Calista Flockhart]. And we would note that it still hasn't generated enough momentum, with fewer than 30 confirmed deals, to keep the company going indefinitely. The question is: Will tightfisted IT buyers open their wallets soon enough for 3PAR to turn the corner?

SAN storage provider XIOtech Corp., which scoots onto the list at No. 10, is actually a restartup.

Founded in 1995, XIOtech was one of the first companies to deliver a SAN-attached storage array. Customers liked the Magnitude storage system – but competitors successfully beat it up because the array didn't provide high-availability features. Rivals also became adept at spreading FUD about the commitment of its parent company, Seagate Technology Inc. (NYSE: STX), which supplies disk drives to most of XIOtech's considerably larger competitors.

XIOtech has now addressed both of these shortcomings. First, it's no longer controlled by Seagate, which sold the majority stake in XIOtech to venture firm Oak Investment Partners (see XIOtech Snatches Rash of Cash). The amount Oak invested in XIOtech hasn't been disclosed, but it's said to be in the neighborhood of $45 million. Ka-chiiiiiing.Second, on the technology front, XIOtech has just released the Magnitude 3D, a modular storage array built on a distributed controller architecture. That design provides not just failover features but also true clustering: XIOtech plans to deliver a 16-node system, with support for up to 256 TBytes of raw capacity, early next year (see XIOtech Fixes a Failing).

Another boost for XIOtech was the recruitment in March of CEO Ken Hendrickson, who brings a wealth of storage industry experience (see XIOtech Hot-Swaps CEO). After 18 years at IBM Corp. (NYSE: IBM), he headed up Ancor Communications, an early leader in the Fibre Channel switch space that was acquired in 2000 by QLogic Corp. (Nasdaq: QLGC).

Hendrickson claims now that Ancor had, hands down, the best technology on the market, delivering an FC fabric switch well before Brocade Communications Systems Inc. (Nasdaq: BRCD). "We had amazing technology, but crappy sales and marketing," he says.

You live and learn. Here's betting XIOtech won't make similar mistakes.

Nishan SystemsPlunging down from No. 2 into the Bit Bucket with disturbing velocity is Nishan Systems Inc.

No, not because it played the role of peevish schoolyard bully in trying to wrest the domain name nishan.com from its U.K. owner, although this episode certainly didn't help its case (see Nishan Bombs Bid for Nishan.com). Mainly, we wonder why a company struggling to gain traction with its products was expending effort on this scurvy exercise. Guys, keep your eyes on the donut!

CEO Bob Russo is appropriately eyeing the bottom line in pushing his sales team hard, and he's envisioning an IPO for the company at some point. But he may also be alienating key members of the Nishan crew, including CTO and VP of engineering Chris Eidler, who quit in April. Since then, Nishan has not named a new VP of engineering (see Nishan Guns for IPO and Nishan CTO Cuts Out).

We also hear that Nishan's backers have been shopping the company around for the past year or so. On its face, that's not unusual. But apparently the only serious interest came from Brocade Communications Systems Inc. (Nasdaq: BRCD), which supposedly made a tentative offer that was far less than the approximately $100 million that the VCs have pumped into Nishan to date. Needless to say, that didn't pan out.

Earlier this year, Russo said he would be looking to raise between $15 million and $20 million in additional funding to tide Nishan over until its putative IPO – but there's been no word on when or whether this funding will come through. The last funding the company received was a $10 million line of credit in December 2001, intended to help Nishan launch its switch (see Nishan Secures $10M Debt Financing). However, an industry source familiar with Nishan notes that while the company is not cash-flow positive, it isn't in danger of imminently going out of business, either.There is some good news for the company: Russo's relentless focus on sales seems to be paying off, with Nishan touting its hundredth customer win – Canadian telecom Telus Corp. (NYSE: TU; Toronto: T) – and claiming it has more than 500 switches in production networks (see Nishan Rings Up 100th Customer).

Nishan representatives, while they wouldn't comment specifically on the company's financial status, insist that it is in good shape. The company's "consistent growth continues to attract investors and keeps us on track to execute to our business plan both this year and [in] 2004," says Nishan spokeswoman Rosie Sundell.

And, as we've noted previously, iSCSI and IP storage networking seem to finally be emerging as a real market. But here's a worrisome trend for Nishan: EMC Corp. (NYSE: EMC) has decided to offer native iSCSI ports directly on its DMX arrays, obviating the need for Nishan's box (see Is EMC Overshooting on iSCSI?). The rest of the SAN market is likely to go the same way, with Fibre Channel on the high end, native iSCSI running over GigE switches everywhere else – and a narrow window of opportunity for Nishan and its brethren.

Will Nishan be able to turn the corner on sales, land enough funding to keep afloat, and keep up development on its next generation of IP SAN switches – while also competing with the likes of Cisco Systems Inc. (Nasdaq: CSCO)? With so many X factors in play, we're moving Nishan off the Top 10 for now.

AppIQAfter a short run, AppIQ Corp. settles into the Bit Bucket with a gentle, powdery poof. We originally gave the startup a spot on the list because it's been at the center of the industry's move to management standards, in the form of the Common Information Model (CIM). Through a combination of luck and foresight, AppIQ was in just the right place at the right time.

But, while AppIQ's software architecture was fortuitously engineered around CIM, it turns out this hasn't been much of a competitive advantage. The chicken-and-egg problem is that, until CIM is widely deployed its usefulness is limited (or nonexistent). In the long run, standards-based interfaces will be a commodity.

That said, things are still looking up for AppIQ. It has plenty of cash in the bank, and there's talk that at least one "major OEM deal" is still in the works for the company, though this hasn't materialized yet. The company also says it has landed six customers, though it's named just three – the Boston Stock Exchange, Nielsen Media Research, and The Weather Channel (see AppIQ Ships 2.0, Lands Six Deals). It's nice (slow) progress, but until it demonstrates greater vitality than the other companies on the Top 10, AppIQ is off the list.

Click on the names of the other companies listed below to read about their fall from the Top 10.

Table 2: The Bit Bucket