XKL Brings New Life To Dark Fiber
New system from Cisco co-founder Len Bosack taps existing cable for low-cost, no-fuss optical networks.
April 17, 2008
![Network Computing logo Network Computing logo](https://eu-images.contentstack.com/v3/assets/bltde8121fc52c5c8f3/blt3f3d0318f746b1c2/65a530e4187606040a1d8b8c/placeholder.png?width=1280&auto=webp&quality=95&format=jpg&disable=upscale)
THE UPSHOT |
---|
CLAIM: XKL promises to revolutionize enterprise networking with an inexpensive, low-maintenance, easily configured gateway that employs wavelength division multiplexing over fiber cable. Its rack-mounted switch provides gobs of bandwidth at a fraction of the cost of carrier-provided optical systems.CONTEXT: The market for optical fiber networks will grow 14% annually over the next few years, according to research firm Infonetics. While there are other niche players out there, the networking industry's leading manufacturers--including Alcatel-Lucent, Cisco, and Nortel--haven't aimed directly at this part of the enterprise market.CREDIBILITY: XKL CEO Len Bosack helped create the enterprise networking market 24 years ago as co-founder of Cisco, giving him bona fides that are impossible to ignore. At the same time, 17 years is a long time to be in startup mode, and the long period of relative invisibility means that two or three generations of networking technology have come and gone since Bosack last made an impact on the industry. |
Networking specialist XKL, a 17-year-old "startup" headed by Cisco Systems co-founder Len Bosack, is getting attention because of that lineage--as well as for the audacious claims it makes for its new enterprise fiber optic switch.
XKL's mission is to let businesses create high-speed optical networks at a fraction of the cost of carrier-provided services. Its DXM Optical Transport System can "change the way" companies deploy IT, says Bosack.
How so? XKL's technology adds wavelength division multiplexing--which creates multiple data channels, each with its own wavelength, over a fiber optic link--to wide area networks. It does so using dark fiber--cable that's in the ground but not being used--at a fraction of the cost of what big carriers charge. "People in the party days were putting fiber in the ground at an incredible pace," says Bosack, referring to the dot-com era of the 1990s, "and there it still is."
Cutting costs is only half of XKL's proposition; the other half is ease of deployment and ongoing management. Optical switching systems tend to require deep expertise, but XKL claims its system can be managed by network administrators who aren't optical networking specialists.
This is where the story of Bosack, an IT pioneer who dropped off the map when he left Cisco in 1990, takes an Oedipal twist. He's out to revolutionize enterprise networking, which means taking on the market leaders, including Cisco, as well as the big carriers. Given Bosack's background, some big companies already are eyeing XKL's system. The company's director of business development, Robert Michaels, won't disclose potential customers but says they're "household names."
The unanswered question is whether they'll be willing to drop their existing carrier-provided networks in favor of equipment from a company that's basically been in stealth mode for a decade and a half.
THE ROAD AHEAD
Driving his 15-year-old Land Rover across the Route 520 bridge into Seattle, Bosack describes the nature of the challenge XKL is addressing. "These fellows--the AlcaLus, the Ciscos, the Nortels--have made optics into a high priesthood," he says. "And it doesn't need to be."
We're heading into the city to meet with David Sinn, chief network engineer in the computing and communications office at the University of Washington, and members of his team. UW is using XKL gear to build high-speed links not just on its Seattle campus but across the Pacific Northwest and as far away as Chicago. It's the first test case of XKL's proposition: a do-it-yourself system connected inexpensively over dark fiber and without proprietary protocols requiring specialized knowledge.
Sinn calls XKL's system "a network engineer's dream DWDM [dense wavelength division multiplexing] box. It's plug and play, it's relatively inexpensive, and you don't have to think about how to make it do what you want it to do."
DIG DEEPER
Time To Upgrade?
You may want to shop around if your goal is cutting-edge network capabilities. We show you how.
Download this InformationWeek Report
>> See all our Reports <<
When Bosack left Cisco in 1990, he signed a noncompete agreement that kept the fledgling XKL from doing anything in the network market for several years. The company started out making mainframe computers that were replacements for Digital Equipment System 20s (Bosack actually helped build those, too, in the mid-1970s). By 2004, "we discovered that the optics industry was starting to make good" on its promises of low-priced, high-volume components, Michaels says.
That led to a fundamental insight: You shouldn't have to have a Ph.D. in optical physics to run a WDM fiber network, says Bosack. Implementing that idea took another three years, what Michaels refers to as the company's pure research phase.
XKL has survived mainly off Bosack's personal investment. Revenue has started coming in, but Bosack declines to comment on XKL's financial situation or its timeline to profitability. XKL isn't Bosack's only professional commitment; he also manages a group of hedge funds.
NO FUSS
The core of the XKL system is the DXM Transport Terminal, a single rack unit for server racks. Designed for efficiency, ease of use, and reliability, the terminals provide Layer 1 WDM connectivity and path protection at bandwidths of up to 100 GBps. Each terminal has 10 channels, and IT departments can use a DXM "band combiner" to stack up to four boxes for a total of 40 channels. A typical XKL system costs about $100,000, or around one-third the list price of comparable telcom equipment from Cisco and Nortel, Michaels says.
DXM uses a conventional command line interface that most IT managers will find familiar. To configure and maintain the system, Michaels says, "all you have to do is type at it."
XKL has a limited track record, but the design point is to create high-speed networks that keep operating with minimal fuss. "Continuous operation is harder than people think," says Bosack. "It should be possible to take this system and run it for years in real time." So far, so good at the University of Washington. Once installed, says chief network engineer Sinn, "I haven't had to touch it again."
Cisco co-founder Bosack is back, ready to simplify WDM fiber networksPhoto by Andy Reynolds |
The XKL system, in theory, should pay for itself. In contrast to a managed DS3 connection from a carrier with recurring monthly charges, it provides bandwidth at a per-megabit cost that declines as usage increases.
The DXM has limitations in terms of geographical range and functionality. The XKL gear lacks "carrier-class attributes," such as fine-tuned monitoring capabilities that carriers demand, notes Ron Kline, research director for optical networks at Ovum, a telecom research firm.
In terms of range, Michaels says the DXM will cover around 100 kilometers without amplification, and with amplifiers spliced in along the fiber route, up to 450 kilometers. That makes it good for metropolitan area and campus deployments, but less so for networking across long distances.
The optical network market last year was $13.8 billion, according to research firm Infonetics. While the bulk of that money came from telecom service providers, 14%, or $1.9 billion, was spent by government agencies, universities and research institutions, financial services firms, electrical utilities, and other businesses. "Most of what enterprises buy is metro WDM gear, exactly what Len Bosack is building," says Infonetics analyst Michael Howard.
RIVALS AND REALITIES
XKL's main competitors include specialists such as Adva Optical Networking, which has sold WDM systems to 10,000 customers. Adva, based in Germany, lost $19.1 million in its most recent quarter.
Cisco and Nortel sell optical switches for enterprise networking, but they're generally more expensive and don't come with the same ease-of-use promises. "We suspect the reason no one has built a system similar to the DXM is because they don't perceive a market or the market is too small," says Michaels.
A complicating factor is that XKL doesn't provide the fiber needed to string its systems together. Bosack may be right that there's plenty of unused fiber in the ground, but gaining access to it can require dealing with brokers, such as AboveNet, or the telcos themselves. Some businesses and other organizations--the University of Washington among them--have rights to dark fiber or their own in-ground networks, but many don't.
Providing fiber access, says Michaels, is something that XKL has considered as part of its future road map. For now, the company can supply contacts and leads to several fiber brokers, such as Broadband Asset Strategies, though it has no formal arrangement with any specific provider. "We haven't seen the need yet" for offering fiber access as part of XKL's service, says Michaels.
If XKL succeeds, it will certainly make some noise: The founder of Cisco, which essentially created and continues to dominate the enterprise networking market, coming back with another game-changing technology for business networks. It's a goal that motivates Bosack.
"Look, Verizon and AT&T and the other big carriers are not going to change," he says, donning his shades for the drive back across Lake Washington. "The optics industry is like the telegraph in the early days of FM radio. It's time for a change."
Contintue to the sidebar:
Where Options Are Few, XKL Finds A Niche
About the Author
You May Also Like
Maximizing cloud potential: Building and operating an effective Cloud Center of Excellence (CCoE)
September 10, 2024Radical Automation of ITSM
September 19, 2024Unleash the power of the browser to secure any device in minutes
September 24, 2024