News Stay informed about the latest enterprise technology news and product updates.

Executive Insight: Allegro poised to disrupt status quo

David Ginsburg studied electrical engineering at Rensselaer Polytechnical Institute. But an equal part of his early computer education came as a member of the Army Signal Corps in Europe, where he worked on a project designed to bring the Internet as close to the front lines as possible. For the past eight months, he has been the vice president of marketing for San Jose-based Allegro Networks Inc., which is currently developing a new router aimed at allowing service providers to set up services more quickly and lower the cost of operations even as the demand for bandwidth grows. Allegro has attracted star power in the name of CEO David House, who was formerly the CEO at Bay Networks. It has also attracted investors, who have kicked in $24 million in first round financing. Ginsburg plans to play a key role in the next phase -- making the company profitable.

The quote on the home page of your Web site reads, ".... winning companies will be those building equipment to cannibalize and disrupt the existing public networking architecture, not sustain it." Is this something of a mantra around Allegro and could you elaborate on what it means?
What is meant by disruptive is that since people have been involved in the Internet it's been business as usual. Providers have been putting in infrastructure and evolving current business models, be they voice or data, and they have been pretty satisfied with trying to eke incremental improvement out of the network. Of course sometimes there have been new technologies that made a (real) difference. The advent of voice over IP is one of those changes. The World Wide Web is probably the best example. But for the most part companies are approaching the marketplace saying, "We have this capacity and in doing this we can help you get a couple more gross margins point out of your business model or these many more ports or this much more bandwidth." Our contention is that incremental improvement works if the overall business model works. But once there is an economic contraction and people start taking a closer look at what makes them profitable, service providers are going to be looking for vendors who can bring a proposition that results in a new business model and a new way of interacting with their customers. Even though you are still in stealth mode, can you tell us more about what differentiates your router?
Service providers are concerned about their T1 businesses. They are seeing more and more commoditization and price pressure and they want to have new ways of servicing these customers, which are typically small and medium business. Our platform allows the carrier to offer wholesale routing services to these ISPs and ASPs so that they can now get new customers in new metropolitan areas without having to build out their infrastructure. For the ISP there is increased profitability and there is a new high-margin revenue stream for the facilities-based provider. It's a win-win value proposition. What are the benefits to service providers and the enterprise?
It's immediately understood that this is something that could finally cause this network expense vs. revenues curve, which is diverging, to converge. They see a potential differentiator that will possibly take them out of this seeming downward spiral of lack of profitability. We've done some business analysis that shows this actually reverses that spiral of non-profitability. What we are bringing into the carrier space is a trend seen all across corporations. It's a trend toward outsourcing and focusing on core competencies and then becoming that much more successful at what your core business model is. We are telling them to focus on their core competency and outsource what they don't need to do So how exactly will your routers differ from, say, something that Nortel Networks or Cisco Systems offers?
Every other device that exists is basically constructed in same way. There's a route processor, switch fabric, and line cards. Every high-performance router is constructed that way. It's still a single route processor sending information to the line card. So that means with a traditional router there aren't dedicated resources and a carrier can't go to a customer and say, "This is your router, this is your buffering, etc." Ours is the first one with multiple route processors. It makes all the difference in a carrier being able to say to multiple customers, "I'm offering you, Boeing, managed services and no matter what happens in the Boeing backbone it's never going to affect Ford's traffic and vice versa. This is a service level guarantee that was impossible to offer in the past. So it's that physical partitioning that's unique. At least one analyst predicts that the so-called edge infrastructure market could soon reach $25 billion. What do you think accounts for the phenomenal expected growth in this area and do you think that estimate is accurate?
That report was from RHK and they were looking at the whole edge infrastructure market. We've also commissioned a report from Gartner and we asked them to break that down further to the outsourced edge infrastructure market. And we came up with an extremely healthy number there, in the order of $7 billion to $9 billion a couple of years out. The reason this market in general is exploding is two-fold. First, we've been doing a good job of building out broadband access, cable modems, DSL, etc. There are literally terabits of new capacity coming online all the time. The big problem is, you've still got this crunch at the edge. You've got more and more bandwidth in the last mile. There is only one way to crack the problem, and that's with an electrically based router. That massive investment has to take place at the edge. The question is, why are there so many more customers? With the growth of DSL and cable there are more and more high-bandwidth customers that want specialized content and we are finally seeing the implementation of viable content models where people get paid for their content. At the same time, we are also seeing immense growth in the bandwidth requirements in the B-to-B space where companies are conducting electronic commerce with one another and that will continue to drive bandwidth. Is the future of broadband bright?
It's like The Six Million Dollar Man. We've got the technology; we can build the network. Now the question is, what do we do with it? What I mean is, what do the consumers and businesses want from the network in terms of content and services, who is going to deliver monetizable content and who along this delivery chain is going to profit from it? There has actually been a slowdown in Internet penetration. People are starting to wonder what they can do now beyond e-mail and a little bit of Web surfing. Content distributors have to have a way of securely injecting content into the network and the service provider has to have a secure way of delivering it to the customers. It's really in creating these end-to-end profitable business relationships that will result in consumers and businesses seeing real utility from the network. There is continual growth of the backbone. The whole question is, what is the user experience? What roles will DiffServ and MPLS play?
I look at DiffServ as part of traffic management and MPLS as partly traffic management and partly VPN-type architecture. At the end of the day, DiffServ is useful because it helps providers at the edge or the customer to identify what types of traffic should get priority over the network. This can help support some of the media applications, video or voice, or ERP, if a provider's network is suitably equipped to handle that marking. And I look at MPLS VPNs as a good way to map customers at the edge of the network across a carrier backbone. Neither one is a be-all or end-all approach. We see them as one way to solve a couple of problems and obviously, since we are working at the edge, we need to support the MPLS and DiffServ markings. Tell us a little about the Broadband Content Delivery Forum, of which you're the chairman.
We're not like a typical standards body. They are very much technology and infrastructure driven. We take and end-to-end perspective and bring together content developers, network services providers, desktop vendors, to not only look at the technology but the business models that hold everything together. Some of the problems the industry is experiencing are due to too much emphasis on basic technology and not enough emphasis on tying it together. It was almost a technological orgy we saw in terms of new hardware platforms being created and everyone trying to outdo each other in terms of the baddest and fastest infrastructure. But sometime in the last six months they woke up and said, "Okay, what do we do now?" One of the problems is taking this to the end user so that they are willing to pay for this infrastructure investment. We need to bring the players to the table to distill the technology into a meaningful business model. I've borrowed a quote from Network Associates. They say, "Who watches the people who watch your network?" Our question is, "Who watches the people who build and deploy the technology?" I just don't think there has been enough of an industry-wide reality check. Last question. What did you learn in the Army that continues to serve you well today?
I think there is an adaptability. You didn't always have everything at your disposal that you needed, and things pop up at last minute and you have to be capable of dealing with changed circumstances. You might find out that you've got to do this thing or that at the other end of Europe on 24-hours' notice. That ability to jump on a moment's notice is important in the Internet environment and is especially important in dealing with customers. Also, you want to look professional at all times. This is especially critical when dealing with carriers. It's not like we're selling a $50 consumer electronics product. The BellSouth's and the WorldCom's of the world are looking for a degree of professionalism that I try to put across at all times. How soon are we going to see Allegro products on the market?
We have the critical mass in terms of engineering and we are beginning to build our whole customer support organization. The year 2001 is more for validation of the platform. We are going to be going to our customers with a beta system and let them kick the tires and test the business models. Our first formal release of the product will be in the first part of 2002. It's an aggressive cycle given the complexity of the platform that we are building but we have made all our milestones, which is a good sign.

Dig Deeper on Network Infrastructure

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.