Baby Steps for a Would-Be Storage Giant

By Peter Burrows BlueArc Corp. may have been the most overhyped startup in the storage market. Launched in 1998, it quickly raised $60 million, hired a big-name CEO in former Compaq Senior Vice-President Enrico Pesatori, and built an innovative technology.

Since the experts figured a 45% growth rate for the high-end storage market, Pesatori pursued a bold growth plan. And when in February, 2000, tech pundit George Gilder wrote that BlueArc "breaks through the storage bottleneck on the Web," many analysts believed that it would soon challenge storage leaders like Compaq, Network Appliance, and even EMC.

Well, not quite yet. BlueArc's ambitious plans were scuttled by the tech spending slowdown, like so many other dreams of overly optimistic startups. Now, analysts forecast 15% growth, and Pesatori has slashed BlueArc's marketing, sales, and hiring plans, and postponed overseas expansion. "We're going to manage top-line growth with the need to conserve cash," says Pesatori.

PLENTY OF QUESTIONS. Fortunately, he saw the venture-capital spigot shutting last spring and raised an additional $72 million in May, rather than the $40 million he had initially aimed for. That should be enough for the company to reach break-even in mid-2003, Pesatori says. By then, he hopes to have revenues of $15 million a quarter.

Thanks to its cash hoard and popularity among digital media companies and government research labs, BlueArc has managed to survive a brutal year. But plenty of questions remain about its future. Analysts wonder if it can gain traction outside of these niches. And BlueArc's engineers may have come up with their breakthrough storage box at the wrong time.

By the time companies cranking up tech spending again, BlueArc may be leapfrogged by a new crop of startups with names such as Zambeel, LeftHand Networks, and 3PARData Inc. "They're a one-trick pony," snipes Network Appliance CEO Daniel J. Warmenhoven. "I don't think BlueArc is going to make it, personally."

FAT PIPE. Dubbed the SiliconServer, BlueArc's "pony" is a new kind of server that's dedicated solely to the job of shunting data back and forth to attached disk drives. That's different from general-purpose servers made by Sun, IBM, and others that constantly juggle multiple tasks such as calling up applications or saving data.

BlueArc was also designed with a far bigger data pipe, handling 2,000 megabits per second. General-purpose machines can digest information at the rate of only a few hundred megabits per second, failing to take advantage of the huge increases in bandwidth made possible by fiber-optic networks in recent years. The result: Even if a company spent gobs on a blazing-fast fiber network, the traffic would clog up.

"The design and philospohy of BlueArc's machine is exactly what I've been looking for for years," says Daniel S. Rosen, head of computer services for Cinesite. A subsidiary of Eastman Kodak, Cinesite needs fast data movement as it handles special effects in films such as X-Men and O Brother Where Art Thou. Otherwise, it's computers -- and employees -- would sit idle for most of the day waiting for the next digital scene to arrive from the storage banks.

EASY UPGRADES. SiliconServer's design could help BlueArc stay on the cutting edge. The functionality of most general-purpose servers is built into their software. But BlueArc embedded its code into chips called field-programmable gate arrays. While more expensive, the code can be invoked much more quickly, since the chips don't have to look for it. And since these chips are programmable, BlueArc's machines can be upgraded with new software -- just like your PC gets fixes for Windows. This fall, Pesatori says the company's 100 engineers will upgrade the product and unveil a new redesign in 2003.

Until then, BlueArc hopes to ride out the downturn by focusing on markets where the storage bottleneck is a major headache. Its product is used in many of the federal government's National Laboratories. Lawrence Livermore National Lab, for example, is using SiliconServer in one of its giant supercomputers, which is used to simulate nuclear blasts.

Then there's the digital media houses such as Cinesite. Throw in a few corporate accounts such as JP Morgan Chase and Nationwide U.K., and the company has installed roughly 60 machines since it began shipping last June. And selling to the customers with the greatest needs creates a windfall as well: They're opting for fully loaded packages that cost around $300,000, far more than the company expected.

MISSED OPPORTUNITY? Some customers appreciate the extra attention they also receive. "The service and support I get out of BlueArc is by far the best I've ever seen," enthuses Cinesite's Rosen. "The only thing I've ever seen that operates this way is oven cleaner."

That's good news, but will it be enough to launch BlueArc into the upper echelon of tech companies? Pesatori says the opportunity still exists. Maybe so, but his chance to make BlueArc a household name may have passed. While he's taking the pragmatic path by backing off its grand plans, some analysts worry that the company doesn't have that luxury.

"BlueArc has definitely built a better mousterap. But you have to take advantage when you're king of the hill. And we haven't seen them dominating as we'd like to see," says Steve Duplessie, analyst with the Enterprise Storage Group. "When the next wave of [tech spending] hits, they're going to need another answer."

By then, another round of overhyped startups will likely be poised with an answer of their own. Burrows writes about technology from BusinessWeek's Silicon Valley bureau

Monsanto vs. GMO Haters

(enter your email)
(enter up to 5 email addresses, separated by commas)

Max 250 characters

blog comments powered by Disqus