Flexibility sets Yipes apart, CEO says
|
|
|||
|
|
Advertisement: |
Metropolitan Ethernet provider Yipes Communications offers gobs of bandwidth to companies looking to connect LANs across a city or the country. Founded 18 months ago and operating in 20 cities, Yipes boasts that it can offer bandwidth at less than half the price of a comparable service from a traditional carrier. Yipes CEO Jerry Parrick recently explained how the company's business model works in a conversation with Network World's Michael Martin.
What lets Yipes offer large amounts of bandwidth at a relatively low cost?
It stems from the breakthrough nature of the architecture we're using. Gigabit Ethernet gives us the ability to satisfy one-megabit customers at the same time that we can satisfy gigabit customers. There is no other architecture that has that range of scalability or that range of flexibility.
On top of that, we can give customers the ability to control their own networking by giving them a Web interface and allowing them to do prioritization and bandwidth online, [in] real time. Moving Gigabit Ethernet out into the wide-area and the metropolitan network and turning it into a carrier-class platform is what essentially is the breakthrough.
What are the bandwidth ranges you're offering?
Customers can pick any combination of megabits from one megabit up to a gigabit in one-megabit increments.
And they're able to do that on the fly?
Today, once they're connected, if they want to alter their bandwidth they call us and within a matter of a few hours we can remotely enable more bandwidth. By the first quarter of next year, we'll have a Web interface where they can go online and change their bandwidth and the prioritization of their traffic.
Our objective is to give them greater control than they've been able to achieve with other suppliers. For example, we're working to allow them to reserve additional bandwidth at a fixed point in time in the future and to have the additional bandwidth added and taken away after the need is satisfied. So if you think of a customer who knows they're going to do a video teleconference a week from now, we want to give them the ability to say, 'Next Tuesday at 10 a.m., I want an additional 4M bit/sec allocated between Point A and Point B, and I want it for four hours.' There's no upgrade or equipment required.
What applications are driving the demand for this bandwidth?
Our core application is built around LANs and connecting them together. So our MAN service connects multiple buildings within a region. The second application we're focused on is connecting LANs to the Internet and providing high-speed Internet access. And we also connect LANs across the country with our WAN service. Any application that runs over a LAN can run over this network. So any application you run in your local computing environment can take advantage of our network.
Who do you view as customers - mostly ASPs, ISPs and large companies?
In our original business model we thought our sweet spot would be midsize, multilocation, regional businesses. We thought connecting five law offices in the San Francisco area would be the prototypical model. What we've found is there are bandwidth needs across the entire enterprise [market].
Who do you view as your competition?
The principal competitor is the incumbent who has all the T-1 and DS-3 business today. And anyone who's saddled with a SONET and ATM infrastructure that makes their services very costly and inflexible.
Can you offer quality of service (QoS), like SONET and ATM, across Ethernet?
Our principal application is LAN connectivity. We're not trying to serve the trading desks at Solomon Smith Barney with 15 tubes sitting in front of every trader and billions of dollars of transactions running every minute. That's not our sweet spot. We may grow up to there some day, but that's not it presently. For connecting LANs, we're very competitive.
With QoS we have queuing capability with four levels of prioritization. By next year we think we'll be competitive QoS-wise with anything ATM has to offer. As for availability, most carriers talk about 99.999% availability. Having been with the traditional carriers for 27 years of my career, I can suggest it's difficult for any carrier to demonstrate they're offering 99.999% availability. We can deliver at least 99.99%, and we're confident that next year we'll be operating at 99.999%.
With SONET, people talk about survivability - if the loop is cut, traffic will run in the other direction. That failover takes about 50 milliseconds. We bring a fiber pair in from one direction and terminate it on one Gigabit Ethernet port, bring the other end of the fiber ring in from the other direction and terminate it on a separate port and that allows the router to work very much like SONET. The difference is the speed of the failover. It takes a few seconds for us.
RELATED LINKS
New optical Gigabit Ethernet switches trigger IP-based MAN data services. The Edge, 12/4/00.
