Sponsored By

PacketFront: Too Good to be True?PacketFront: Too Good to be True?

With just 36 staff and a few million dollars, the startup has made amazing progress in broadband access equipment

October 9, 2002

5 Min Read
PacketFront: Too Good to be True?

Once in a while, a story crops up that sounds too good to be true. This is one of them, and it concerns PacketFront AB, a Swedish startup developing broadband access equipment that today issued a couple of press releases (see PacketFront Fires Up Broadband).

Yep, the press releases are pretty dull. But PacketFront isn't. For a kickoff, it only got its first funding in August 2001, and yet its CEO, Martin Thunman, expects to go into profit next year.

PacketFront already has contracts with "five to ten" customers in Sweden, and if those customers follow their rollout plans, PacketFront's revenues will total $30 million in 2003, $60 million in 2004, and $100 million in 2005, according to Thunman.

And that's just for starters. PacketFront is in final negotiations to supply its gear to another 20 operators and is talking to a further 100, according to Thunman. "We don't have any problems getting to talk [to potential customers] because we're so different," he says, adding that PacketFront has an 80 percent win rate with prospects.

So, how is PacketFront "so different"?

First off, PacketFront is addressing the market for genuine broadband residential services -- capable of handling multiple TV channels, games, telephony, and so on; where each home requires a minimum bandwidth of 25 Mbit/s and has to be connected with fiber, not copper.

Existing technologies for delivering services over DSL (digital subscriber line) connections simply won't scale to these bandwidths, according to Fredrik Swartz, CTO of ViaEuropa, an operator that's using PacketFront's gear to deliver broadband services to more than 10,000 households in Sweden.

Swartz has discovered this the hard way. Initially, ViaEuropa installed equipment from Dynarc, a Swedish startup that went bust earlier this year. "The bloody Dynarc things didn't work properly. We couldn't bill anyone," says Swartz. ViaEuropa had to install Ethernet switches from Hewlett-Packard Co. (NYSE: HPQ) as a stopgap measure, knowing that it would need a Layer 3 network in the long run.

Why was a Layer 3 network so important? ViaEuropa is pioneering a business model that's beginning to gain traction around the world. The model envisions the service provider market dividing into three layers. The bottom layer -- passive infrastructure like fiber and ducts -- is often supplied by local governments, utilities, or real estate companies. The next layer up -- where ViaEuropa plays -- is the active infrastructure over which players in the third layer can offer services to end users. The end users themselves aren't aware of ViaEuropa, but as users are likely to buy lots of different services from lots of different providers, ViaEuropa needs to use routers rather than Layer 2 switches to handle their requirements efficiently.

A Layer 3 solution is also necessary to support multicasting, a prerequisite for running broadcast TV over broadband networks, adds Swartz.

After its early experiences, ViaEuropa took a long hard look at Layer 3 architectures before picking PacketFront. It reviewed a proposal from Nortel Networks Corp. (NYSE/Toronto: NT) for a "service node" scheme similar to the ones in widespread use in DSL deployments, only to conclude that its centralized architecture wouldn't scale.

PacketFront avoids this centralized approach with its scheme, which comprises access routers installed at end-user sites, controlled by what's called a Broadband Ethernet Control System (BECS) at a carrier site.

The key point is that the BECS merely controls the configuration of the access routers. It doesn't handle the actual traffic going between the access router and the provider of the particular service, as is the case with DSL service nodes. As a result, it isn't a potential bottleneck.

Swartz notes that the BECS approach has other advantages as well. Installing or replacing an access router is a breeze, because it's automatically configured over the network.

PacketFront has a third product, a "Service Provisioning Ethernet Control System" (SPECS), that decouples the provisioning of services from the configuration of the network -- a prerequisite for offering end users a free choice of services while masking a lot of complexities from operators.

PacketFront's access routers also deserve a mention. They come in 8-, 16-, or 24-port versions and are purpose designed, which means there's no need for extra fiber optic conversion equipment; and they don't need cooling fans. They also boast 64 Mbytes of packet buffering, much more than in the average access router, according to Thunman. All of this equates to a "cost effective" solution, says Swartz.

For the record, PacketFront and ViaEuropa are totally independent. They haven't invested in each other, nor do they share investors. ViaEuropa has been using PacketFront's equipment in its network since May and is happy with it. The BECS has had one outage of three hours in that period, and as a result, ViaEuropa is planning to install a backup. It expects to be rolling out 2,000 ports this year and to have bumped that up to 20,000 in 12 months. That equals a lot of PacketFront equipment.

PacketFront appears to have achieved an incredible amount, bearing in mind that it's only got 36 on its staff and was founded only 18 months ago. Thunman puts some of this success down to the experience his staff gained at Bredbandsbolaget AB (B2), another startup operator rolling out broadband services in Sweden (see Swedish Carrier Suffers Setback). Most of PacketFront's staff either worked for B2 or worked for Cisco Systems Inc. (Nasdaq: CSCO), which supplied much of the equipment for the B2 project.

PacketFront hasn't disclosed how much money it raised in its first funding round, in October 2001 (in other words, it was embarrassingly small), but it only raised SEK43 million (US$4.6 million) in its second round, announced last May. Investors include the U.K.'s TLcom Capital Partners Ltd. and European Equity Partners. Chairman of the board is Stefan Lindeberg, a former Cisco employee who now has stakes in a number of Swedish startups.

All of this begs a big question: How can 36 people with a few million dollars achieve so much in so little time? If it wasn't for the fact that ViaEuropa is actually using PacketFront's equipment, it definitely would sound too good to be true.

— Peter Heywood, Founding Editor, Light Reading
www.lightreading.comWant to know more? The big cheeses of the optical networking industry will be discussing broadband access equipment at Lightspeed Europe. Check it out at Lightspeed Europe 02.

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like