Next generation branch networking

NetD continues to challenge the big guns

SGI logo hardware close-up

As a start-up company your funding typically follows phases corresponding to prototype, product launch, then breakeven/growth. You have to match your "burn rate" to this timetable, and execution on your plan is key to ensure you don't run out of money, or give investors reason to doubt your ability to take the business forward to a sustainable market position.

Introducing a product that seeks to replace an incumbent is easy for investors to understand, but has a small margin for execution error as it can be very hard to differentiate. Procket Networks is one such example: everyone knows what a core router is and does, and after $272m in funding to build an alternative, the company failed to grow revenue fast enough to make investors believe it would displace Cisco or Juniper. Cisco itself eventually picked up some of the pieces for under $90m when assets were sold off in August 2004.

Such failure is a nightmare for customers. Left with installed equipment that has no long-term future, they usually end up replacing it with the incumbent's solution that they chose to avoid in the first place.

Looking for a gap in the market and filling it with a different kind of product is another approach. In the network equipment space, attacking the edges of the network with innovative offerings, either at the central office gateway or the branch office, is a strategy that has been successfully executed. Wireless networking and gateway security vendors such as Fortinet are obvious examples.

The trends in branch and remote office enablement open some new doors here. The ability of branch office staff to access core business applications deployed centrally is key, but increasingly, there is also a need for integrated communications. Productivity enhancers such as IP Telephony and other collaborative applications like calendaring, messaging, conferencing and video, contribute to the need for an open but secure multi-threaded communication between the branch office and the centre.

Looking back at history, this is the latest in a series of waves of branch office automation, from the first migrations of asynchronous terminals to remote LANs, through the rise of remote application access using technologies such as Citrix, to the delivery of converged voice and data networks over a common secure IP infrastructure.

At such moments, there is always the opportunity for a new player to arrive who has a claim to a "clean sheet" approach to the branch office problem. Cue a new class of product called the Business Gateway (by InStat) or Services Gateway (by IDC). These are both ways of describing a new entrant product sitting at the edge of a branch office network that blends security, routing and converged IP communications.

Seeking to take advantage of this shift is Sunnyvale, CA NetDevices (NetD), founded in 2003 by three Cisco veterans. If you wanted to pick a group of people to build a better gateway routing product you could do much worse than a team that worked on the Cisco 7000 series router. But what is it that this relatively new 165+ person company does differently?

NetD's core argument is that its new modular OS and hardware architecture is highly reliable and capable of handling all common branch network functions without running out of processing puff. NetD contrasts this with "traditional" multi-function boxes that it says can suffer from incremental performance degradation as each new function is switched on. It also claims superior remote management capability.

It's a compelling story, but success is just about technology. We need to consider the company's ability to execute, as developing and continuing to evolve such a platform involves a lot of coordinated work from ASIC, hardware and software designers, which in turn needs a lot of VC dollars.

On the surface, it looks like NetD has this covered. The company raised $25m in August 2005, which should keep the engineers writing code for a while. But we have to come back to the old question of whether it's going to be enough to unseat the likes of Cisco and Juniper, who are continuing to develop their branch networking offerings.

We also have to be realistic about the proposition. Although it may be true that branch offices have too many single-use devices (NetD claims up to six), you will never reduce them down to one. NetD, for example is missing some key branch office security functionality such as URL and SPAM filtering, as well as gateway anti-virus capability. It also doesn't include wireless management, which is increasingly in demand in the branch office context.

And will buyers really trust devices with such a broad capability? Many enterprise clients increasingly see security as a number one branch office issue, for example, as branches go out to the internet directly rather than via the central office. Others may require functions like email encryption, specific multiplexing or network termination. Equipment choice is therefore going to gravitate back towards best-in-class products in many cases as customers try to avoid perceived compromises. This means the ground continually shifts under players such as NetD to demonstrate that their blended implementations rank alongside the established incumbents. Is NetD's routing really as good as Cisco or Juniper? Is the IDS/IPS as good as Tipping Point? And what about the Firewall?

The reality is that during the market window that NetD is enjoying, the big vendors with their large installed base and big R&D budgets are moving in to plug the gaps. Cisco has added capabilities to its Integrated Services Router (ISR) which holds the centre ground for branch office routing and is the product of choice for many service providers, who are the key to securing volume sales. Juniper introduced its Secure Services Gateway (SSG) in February which comes from its NetScreen acquisition, with very aggressive entry level pricing. While NetD has just announced its new SG-4 product to compete at this same level, the competition is clearly hotting up.

So the key questions regarding NetD's future are:

  • Has it picked the right core features to integrate?
  • Is it able to execute adequately?
  • Can it generate positive cash flow quickly enough to secure the additional funding it is likely to require to move on to the next stage?

At the moment, NetD is off to a good start, but the big incumbents are not going to make life easy for it. We'll therefore have to see whether this innovative challenger can hold its own over the longer term.

David Perry is principal analyst at independent research firm Freeform Dynamics Ltd.

Sponsored: Minds Mastering Machines - Call for papers now open

Biting the hand that feeds IT © 1998–2018