CBR Q: Juniper has had a difficult year with declining revenues, 570 layoffs and other cost cuts. Where do you expect the overall growth to come from next year?
Yes, we are little bit disappointed in the year. Fundamentally, we see two big market drivers for our growth next year. The first is around all of these companies building clouds, including the biggest service providers and enterprises building their own private cloud environments. Some of our recent customer wins like UBS and Nike are saying I want my IT infrastructure to be agile, on-demand and just like the big public cloud providers. Those customers are in an investment cycle right now and will be for some time in building these cloud environments.
The other big chunk of growth is this concept of intelligent networks, which means that the networks that we enable our customers to build can become smarter and have better analytics. If all of this data is being created, the network providers are the ones that are sitting on a gold mine of information that they often can’t tap into. So giving them the ability to better understand their customers and what their customers’ interests are, can improve and customise services.
CBR Q: Your routing product revenue fell 12% year-over-year in your third quarter of 2014. As more companies move to the cloud, will there be a further decline in demand for routers? How will this impact revenue?
If you look at one of the big markets that we serve, telecoms providers, the cloud is basically an extension of the network they have, and so putting more services in the cloud really means putting more services on their network. This may be delivered in a slightly different way, but we still believe it’s going to be a key part of their business.
With the virtual MX router that we announced in November 2014, there are going to be natural laws of physical limits to what the virtual MX can do if it’s running on a generic server that runs X86 chips. This is because x86 can only scale so much. So it’s going to be a nice continuum of smaller applications or early testing of new markets, which allow you not to have a big upfront cost to trial new things. And if some of those things take off, then you can scale to a point where you’re going to need physical instance of MX because we build custom chips to give you massive levels of throughput and performance. So we have now a fairly complimentary set of options: the physical and the virtual.
CBR Q: What differentiates the MX router from Cisco’s, Brocade’s and other providers’ virtualised routers?
Brocade is fairly new to the router game and so the level of features and capabilities that they have are very basic. If you think about how service providers use routers today, they serve business customers and business customers need the ability to offer VPN services and security services, which are not built into the Brocade solution.
For example, if information comes through the router and if all I need to decide is what is the next place I need to send that information, that’s the most basic part. If instead, I have to figure out what’s in this packet; is it video traffic? Do I need to run it through a firewall to make sure that it’s not bad traffic? That’s many more things that I have to think about, so can I still do 100 per second, and that’s the question. So with Brocade, they don’t even have all of those functions. They can’t even do all the other things and if they could it would bog them down….They need functions that support subscriber billing and monitoring and three screen delivery to the home and all of these functions that are router performed. These are again things that we have invested in for years that Brocade doesn’t have.
Cisco has two different offerings running two different operating systems and we lead them in feature capabilities. We also lead them in scale and performance. We’ve a much more comprehensive feature set but then we also have the level of scale. This means that the amount of traffic that the virtual router can process, ours is higher than anything we’ve seen from Cisco. We’ve tested them in a lab. I don’t know what they claim but in actual real testing we significantly outperform them in how much traffic we can pass through.
CBR Q: You also introduced Contrail Cloud in November 2014. How does it differ from other offerings?
Contrail is our SDN or network functions virtualisation solution and what we’ve done is tested not just our own solution but also some of the other vendors’ so we have complete solution sets that we can hand to our partners to assemble and deliver to our customers. We’ve done something slightly different which is if you go to Cisco or HP to get a similar stack solution, they’re going to give it to you fairly hard wired, meaning it’s got to be like this and for Cisco it’s got to be all Cisco stuff.
Rather, we’re going to give you the blueprint of a number of different options and then a partner. For example, Intec takes that blueprint and they can sit down with the customer and understand what their current environment is and what they need, and then pick the right pieces that are already pre-validated to work together, but the partner gets to assemble that, and that’s the value added position we want our partners to be in because they’re the experts in system integration.
CBR Q: What’s your view on the open vs. proprietary SDN battle?
We really want to provide as many options as possible because the world is changing very quickly and I think customers are becoming wearier of locking in with one vendor and assuming that that vendor is always going to be the exact right solution for them.
Whereas 10 years ago, I could have maybe said well I’m just working with that one company now because of how connected we all are because the network is such an important delivery vehicle for insights and applications, now almost every company and who they’re dealing with is increasingly important to them. They’re looking for a new level of agility and performance from the network that many of them never really cared or thought about.
It’s a more sophisticated step than an average enterprise, so these people really understand that the network and IT in general is a differentiator for their business and so they can’t afford blind lock in any more. They’ve realised that and so they’re willing to really look at the different options that are out there and they want to keep their options open.
CBR Q:Juniper made recent cost and job cuts. What are the effects so far?
The biggest impact has been focused on the areas that we believe are high growth and high opportunity for us. And that’s meant a little bit of refinement of our product portfolio so we announced the sale of one of our products called Junos Pulse. We have made a few other minor tweaks to our roadmap and cancelled a few projects that hadn’t been launched yet with the goal of driving focus on these specific industries, the specific types of customers and the places where we think we’ve got the best opportunity to win.
CBR Q: Are you looking to get more people on board Juniper with new skills or are you looking to get rid of more people?
In these new investment areas, we are constantly bringing in people with new skills and capabilities around software defined networking and the key focus areas. We need to make sure that we have strong capabilities around the cloud, security and routing, so we’re always looking to broaden our skills sets. We don’t have any plans for any kind of digital companywide reductions but we’ll always look at the portfolio to see if there’s an area that we say we want to do well, but it’s not with the goal of reducing overall company headcount.