Bernd Schumacher_ NSN Keynote

Document Sample
Bernd Schumacher_ NSN Keynote Powered By Docstoc

                    2008 EUROPEAN PRESS SUMMIT

                               Unedited First DRAFT

                                 Keynote Presentation

                                       Bernd Schumacher
                     Head of IP Transport, Nokia Siemens Networks

     The world connected in 2015. We believe that 5 billion people will be online, will be
     connected by 2015. This is a massive uptake on the subscriber side, on one side, and
     massive uptake on bandwidth. How will this be possible?
     On one side we have access technologies, broadband access technologies that will
     drive this. Whether this is wireless technologies or whether this is wireline
     technologies, broadband will be everywhere. Broadband services everywhere, it's the
     paradigm technology that's giving us today already this possibility, mobility as well as
     on the fixed network side.
     The applications will be predominantly in the Internet. This in itself will drive a
     massive demand for bandwidth. Centralised application distributed in the Internet
     will drive massive demand for bandwidth services and, combine this with new
     business models, would have to evolve because, on the operator side, the value chain
     will change and they have to align value creation along a new and distributed value
     chain, will also drive massive demand for bandwidth.
     This vision is really in the heart and is in the centre of our strategy in terms of moving
     forward and how to evolve transport architectures into next-generation networks. Let
     me have a first look into some studies in how the bandwidth uptake is expected to
     On the left-hand side you can see predictions of how the bandwidth demand on a
     household level will grow over time. And you will see basically three prediction
     levels, 2010, 2020 and 2030. Obviously when you go to 2030 there's a crystal ball.
     But 2010 is relatively close ahead.
     And you can see that there are a variety of predictions, here from the lower end up to
     the higher end. But we expect already, let's say, on the higher-end prediction that 100

Quinta do Lago, 25-26 September 2008
2008 European Press Summit                                                                   NetEvents

     megabit per second per household will be required to serve new services by 2010.
     This is the higher-end prediction. You may say okay, this is not coming through. It's
     the higher end. But if you balance this with our vision in 2015, 5 billion people
     always on and always connected, we will see that somewhere in this timeframe
     between 2010 and 2015 it will most likely happen. And we will need 100 megabit per
     second down to the household, driven by several different services.
     On the right-hand side you can see what does this mean in terms of bandwidth uptake.
     It is really the exponential growth that we have to anticipate and we have to expect.
     This right-hand side is a study we called 100 fold traffic increase study that we have
     done internally, with a lot of operators globally. And it indicates the drivers really
     behind it. On one side it's residential services. Whatever you have in terms of online
     services, high bandwidth services is around video today. And video services will
     really drive on a residential side the bandwidth uptake.
     On the residential side, Internet traffic, it will be also peer-to-peer services, user-
     generated content, if you look to YouTube, photo sharing and whatever, you name it,
     drives it on the residential side.
     But also the enterprise sector plays a quite important role here, the demand for secure
     data networking, data storage and also more centralised IT applications will drive
     bandwidth demand coming out of the enterprise arena. Those factors together will
     lead exactly to this dramatic bandwidth growth. And you might have noted that on
     the scale, we're talking here about 100,000 of petabytes per month. We're losing now
     slowly the units how to measure this. Now we have finally arrived at petabytes per
     month. And this, in itself, tells a story.
     There are significant challenges around this bandwidth growth and tremendous
     opportunities for the industry because we have to manage this bandwidth growth in
     extremely cost-efficient way. On one side, bandwidth is getting a commodity. And
     we have to make sure that we add value beyond pure bit pumping. We will have 5
     billion people. We will have 100-fold traffic increase. And this puts a huge, huge
     pressure on the operators, on the network infrastructure. It is all about driving the
     costs down. It's all about driving the cost down from a total cost of ownership
     At the end of the day, differentiation will come on several parts. First hand, reliability
     of the network infrastructure. This is absolutely a key. Second one I mentioned
     already, cost, cost, cost. The combination of reliability and cost will be absolutely
     key element of differentiation because the operators are under huge pressure. They
     have churn. Customers are running away. They have to implement a new way of
     designing networks very fast.
     If we look into today's network architectures, then today's network architectures are
     not ready for this challenge. We have to make sure that, on one side, let's say,
     traditional ingredients that we see in SDH networks, ATM networks, frame relay
     networks are preserved. This means high availability, resiliency of those networks.
     And on the other hand we have to make sure the data-centric networks will really
     scale to this bandwidth demand. And we have to preserve, at the same time, the

Quinta do Lago, 25-26 September 2008                                                                2
2008 European Press Summit                                                                    NetEvents

     security, the control mechanisms that we are used to out of telecommunication
     industry in those segments, in particular from the network management aspect. We
     have to make sure the quality of service is all over the place, that it can be measured,
     that at the end of the day service level agreements can be tailored exactly according to
     hard-defined QoS parameters.
     This is a study that we done with a significant number of global operators in the '07
     and '08 timeframe. It looks really into the business challenges here that the operators
     face. First of all, competition. This means fear of churn from the customer side, from
     the end customer side. It's revenue growth and it's profitability. And this results
     really in the top challenge that we have today in the industry. We have to enable
     really the service providers to compete with new services. At the same time we have
     to reduce the complexity and have to take the cost out of the system.
     This will be the key element. Whenever I'm meeting my customers today, it's about
     taking cost out, taking cost out, taking cost out, scaleable, very effective solutions, but
     cost is key. That's the key element that we have to drive.
     With the technical challenge, the key posted elements are integration with legacy, so
     sometimes we might tell you that this is the Ethernet revolution. It's actually not
     because we still have to deal with significant legacy networks. And I would say the
     key challenge is to master the boundary line between evolution and revolution. We
     have to take into account significant installed bases obviously. It's rapid provisioning,
     reliability and manageability of the networks.
     Let me take the end-user perspective here, all of us here in the room. Whenever we
     render a new telecommunication service, what do we really care of? I mean once we
     order the service, we want to have the service activated as quickly as possible. This is
     today a significant challenge on the operator side, fast activation, fast provisioning of
     the service. Whenever we have the service, we want to make sure that we have the
     right service experience and reliability of the network. This can come in perceived
     reliability, but it will come more and more in particular on the enterprise side, with
     hard level, service level agreement, QoS parameters, measurable, predictable, and you
     pay for, at the end of the day, what you get. This will be one of the key elements.
     And the third element, cost. It's cost, cost, cost. As said, today's network
     architectures are not fully geared to solve those challenges and we have to find a new
     way to address this.
     We believe that Ethernet and photonics play really in the combination. Ethernet and
     photonics play an absolutely fundamental role in addressing this challenge. It is
     Carrier Ethernet transport. It is Ethernet technology that we all know 99% of the lines
     today of the world are running Ethernet protocol. It's widespread. It's clearly the
     protocol of choice. But this is available today in a much better value proposition in
     carrier-class resiliency, in carrier-class availability, in carrier-class manageability.
     In combination with photonic network architectures, high-capacity networks can be
     built in an extremely cost-efficient way and in an extremely reliable way. These are
     really the two pillars of our strategy, also of our investment strategy, that we invest

Quinta do Lago, 25-26 September 2008                                                                 3
2008 European Press Summit                                                                 NetEvents

     extremely heavily into Ethernet architectures, Carrier Ethernet architectures and
     photonic networks. And this is not only about Ethernet switches and carrier-class
     resilient Ethernet switches. It is about Ethernet, carrier-class Ethernet that will
     penetrate all our traditional transmission products.
     It moves into next-generation SDH system, it's an integral part of microwave
     backhaul systems, with the carriers and switches, of course. But Ethernet, Carrier
     Ethernet ingredients will move and are already moving into DWDM systems, for
     example, because we need the Ethernet aggregation capabilities also on high capacity
     DWDM systems. This is the common theme. That's really the red line how we drive
     our investments and how we drive our network strategy moving forward.
     This might look like a very techy slide. Sometimes you can't avoid obviously the
     technological view on this. But at the end of the day, we have to look in to the
     business part here. And we have to broaden our view actually onto all network layers
     and to optimise networks in a more holistic approach.
     And we have layered networks on each. Layer one service, layer two and layer three
     technologies in the networks today. A further optimisation of the cost part of the
     network can only be done if we flatten the network architectures. This means flatten
     network, i.e. things around Ethernet and optics. And if you take a holistic view onto
     network architectures, this means we have to make sure that services are routed, are
     processed on the lowest possible layer.
     Optical technologies give us today already flexibility to route specific service on the
     optical plane, to switch service on the optical plane. Ethernet gives us the next layer,
     and IP gives us the next layer. In terms of driving cost efficiencies and driving costs
     out of the network, and integral view of the network is necessary. And we call this
     multi-layer optimisation. It makes sure that we process the traffic on the lowest
     possible layer.
     How do we do this? It's obviously not a one-size-fits-all approach. We do this in a
     very heavy consulting process with the leading network operators here globally, by
     looking specifically into their current network architectures and how we can optimise
     basically traffic distribution on those different layers. We're doing this also with
     heavy support of tooling, with heavy support of simulation tools.
     And only to give you a little bit of flavour, what we can optimise really in terms of
     cost in the network architecture, in real life network example of European operators,
     we managed to take 70% cost out of current networks, which is holistic network
     optimisation approach. So it is Ethernet and photonics as key pillars. And it is
     overall network optimisation across all three layers.
     Direct benefits can be realised very quickly by introducing Carrier Ethernet transport.
     And those benefits come in many dimensions. And I will go in a little bit more detail
     into the individual dimensions.
     Ethernet economics. Cost efficiency of Ethernet solutions, scalability of Ethernet
     solutions, but also a complete industrial ecosystem around Ethernet technologies.
     This goes down to component industry, to software modules, that drives overall

Quinta do Lago, 25-26 September 2008                                                              4
2008 European Press Summit                                                                   NetEvents

     Ethernet efficiency and cost. Network operators that have been early adopters, on one
     side on greenfield architectures, on the other side also on fast-evolving architectures,
     have captured on this Ethernet potential already today. And I will come later on to
     specific examples of network operators that have done this, that are in the middle of
     this transition and that can drive really significant cost out of this.
     Connection orientation. This is one of the key ingredients of Carrier Ethernet
     transport. You know, at the end of the day, with a connection-oriented approach,
     where is your service really running over the network. It's about predictability. It's
     about quality of service ingredients. And it's about reliability.
     With connection-oriented approach, we can make sure that we really control on a
     service level, not only on a pure connection level. This is very important for hard
     level quality of service agreement for SLAs. And this will be extremely important
     start obviously today more on the enterprise type and business applications, but this
     will be a common theme. At the end of the day, the consumer will pay for what he
     gets. And therefore a connection-oriented approach is the most efficient way of doing
     A deterministic network behaviour is also extremely important. Whenever there is an
     issue in the network, we have to understand, where is it? It's about full isolation
     mechanisms. It's about pinpointing specific trouble points in the networks and not
     being left out in the cold that something is routed to somewhere and nobody knows
     where is the real issue. This is especially in combination with the drive of end
     customers towards hard quality of service, a key topic.
     Carrier-class resiliency. This is about bringing carrier-class reliability into hardware
     design, into software design, with really the high availability of the networks. But
     we're talking also about fail-over mechanisms, fast restoration, 50 millisecond
     restoration time that, at the end of the day, our customers are used to in, let's call it,
     traditional network designs. Next-generation scalable solutions around Carrier
     Ethernet transport will give exactly and do give today the same value proposition.
     And at the end of the day, this has to be complementary with end-to-end OAM
     procedures because you have to measure really what you get. This is the key to
     provide QoS.
     Full layer 3 transparency, secure transport services, and topology agnostic approach is
     also very important. We consider Carrier Ethernet transport the ideal means of
     transport layer 3 services. And it is about really flexibility of the network design. It
     doesn't matter whether, from a planning perspective, this is ring/mesh, spool,
     whatever time of architecture in a network. This is a topology agnostic approach that
     gives a lot of flexibility really to the operator. And the network planners, they like it
     because the world is not so predictable as we love it to be sometimes, especially not if
     you're planning data high capacity networks. And flexibility in planning is one of the
     key topics that are required.
     I already explained the multi-layer service management approach. It comes down
     with simplicity of management, but it comes with the holistic view onto the networks.

Quinta do Lago, 25-26 September 2008                                                                5
2008 European Press Summit                                                                 NetEvents

     Let me go into some specific examples, because this is not future music. You might
     call it future music, but the future has already started today.
     Example number one of proof point of Carrier Ethernet transport introduction, COLT
     - a leading European network operator, providing business services; a pan-European
     operator; 13 countries in Europe; 35 metropolitan areas served; 25,000 kilometre,
     quite extensive network. So their challenge was really to deliver very innovative
     Ethernet-based services in a very flexible way, in different service aspects, E-Lan, E-
     Tree, and with flexible service attributes, committed information rate, speed,
     scalability and fast change of those service value propositions. This was really their
     key challenge. At the same time, since they are working in the enterprise
     environment, hard-level SLAs are their today bread-and-butter business.
     And the end solution that we have put forward into this network is based on our AC
     [raised] Carrier Ethernet transport switch portfolio. They are rolling this out now all
     over Europe. And it is combined with our best-in-class ASPEN network management
     system to manage really on a very easy to use, point-and-click approach to manage
     Ethernet networks in end-to-end approach, with service level agreement, with point-
     and-click capabilities, like the way the operator always used it in traditional network
     designs, but with the scalability benefit around Carrier Ethernet transport.
     This has resulted in a significant lower total cost of ownership, in particular on the
     OpEx side. It gives them a really hard level QoS for whether this is point-to-
     multipoint services or point-to-point services. 50 millisecond fail-over and restoration
     time, exactly what the customers were used to because they were asking for it as well.
     And it gives them a very nice bridge to a legacy role because they have full support of
     their traditional TDM infrastructure with [circuit innovation]. Only one example of
     the fact that CET is reality today.
     Multimedios, a second example. Number three cable operator in Mexico, providing
     triple-play services, but not only providing triple-play services. At the same time they
     are a carrier's carrier, and they provide mobile backhaul services for all leading
     operators in Mexico. Their challenge was slightly different. So they had to provide,
     on one side, this carrier's carrier approach. On the other side they had to provide a
     means of very effective video distribution.
     Also here, method of choice, introduction of Carrier Ethernet transport, Carrier
     Ethernet transport technology in combination and as an integral part together with our
     photonic solutions in the DWDM space. This all embedded in a comprehensive
     management system, end-to-end managed.
     And they've clearly seen a way of calculating this really down to the penny, the
     savings of Carrier Ethernet transport against traditional layer 2 switching. So it's
     really a challenge of migrating very smoothly this unified network from what they do
     today into also providing same service, like carrier's carrier services over the same
     network architecture. So current network architecture, Carrier Ethernet transport
     based, in combination with photonic networks. And this has proven to be the most
     cost-effective solution for them.

Quinta do Lago, 25-26 September 2008                                                              6
2008 European Press Summit                                                                  NetEvents

     So I hope I've given you a little bit of flavour that this Carrier Ethernet transport
     future has started already today. We see a significant uptake on the bandwidth side.
     We have to change the way, as an industry, as design those networks. We have to
     define and we have to find new and most efficient network designs. This is what we
     do with Carrier Ethernet transport. This is what we drive into the market. This is
     what really has a very high acceptance right now.
     The future has started today. Let's have an interesting discussion around those topics
     today on NetEvents. And I wish all of us really a very productive session here and a
     lot of interaction. Thank you.

                             Interview and Audience Q&A

     Manek Dubash – Editorial Director, NetEvents TV
     Bernd, thank you very much for that. An interesting exposition of the brave new
     world of Carrier Ethernet. But, as we all know, A, it's not going to be easy, and B, it's
     not going to be instant, in spite of there are a few case studies scattered around the
     world. But how long is it all going to take? Some analysts are saying 10, 15, maybe
     20 years.

     Bernd Schumacher
     Yes. Transition cycles in telecom networks, it's always a question how you measure
     this. If you look into different other transitions that we had in telecom networks, you
     might say yes, until the last boxes fade away from the networks, it might take ten
     years. But I think the pressure points are different ones today. We have new services.
     We have new bandwidth-consuming services. We have 5 billion people that will be
     on and will be connected. And the economics have changed.
     Only one example, mobile backhauling, whereas, let's say, in the past, I would say the
     operators didn't care really on the mobile side about backhauling costs, based on
     leased lines they paid a hell of a lot of money. With the uptake of bandwidth on the
     mobile side, with LTE coming, this is now a very, very critical element in the
     business case from a cost element. And simply the pressure to change over to more
     efficient architecture is up. Therefore I expect that this will gain ground very, very
     And we see simply from the interaction with our customers how much concern there
     is around this, how to build the most efficient network architecture. And I predict it is
     moving very fast.

     Manek Dubash
     So from and end-user perspective though it's still going to be the case that if you live
     in a town, if you're one of the many billions of people who live in town, then fine.
     But if you don't, then what?

Quinta do Lago, 25-26 September 2008                                                               7
2008 European Press Summit                                                                    NetEvents

     Bernd Schumacher
     Yes. If you take it from a pure bandwidth volume perspective, I think we all
     understand, it is clear, this start and will kick off more in the large metropolitan areas.
     But it will spread quite fast. And I think especially here also with mobility and
     broadband mobile services, also rural areas will have very soon basically penetration
     of this one.

     Manek Dubash
     Okay. So what you're saying is the whole thing's going to be funded and made
     attractive to people through services, but carriers have really not exactly covered
     themselves in glory when it comes to providing services that consumers actually want.
     It's been third party services providers in the main who've done that. So how are the
     carriers going to get the revenue, because basically they don't have a track record in
     this department?

     Bernd Schumacher
     Yes. That's the multi-million-dollar question. I think all the operators globally, of
     course, they ask themselves exactly this question. But at the same time, one must say
     all the new services that are discussed, all the news services, let's say, that are gaining
     shape, they've one common element, and this is broadband. So whether you take
     video services, IPTV, video-on-demand services, so more on the residential side, or
     whether you take enterprise services, highly reliable networks on the enterprise side, it
     doesn't matter. The common element is they will consume significantly more
     And this is, of course, at the same time the operators' dilemma because we have, on
     one side, this massive uptake on the bandwidth and the revenue doesn't grow
     accordingly. And this is exactly what we're here for. And this is what we have to
     drive, simply more cost-efficient network designs. Bandwidth is a commodity already
     today. It will become even more a commodity. And the willingness to pay for the
     bandwidth itself in the value chain was limited. So this means, simple answer, take
     cost out of the network structures. And this is what Carrier Ethernet transport will

     Manek Dubash
     But how can they protect themselves against basically the revenue being sucked out of
     the system by the third party service providers?

     Bernd Schumacher
     Yes. Revenues being sucked out of the system, first of all it's differentiation. It is
     differentiation, and exactly the ingredients that I mentioned. It is really reliability of
     the network, best-in-class cost position of this network, end-to-end manageability of
     those networks. And it is best service, end-user service experience. And this is the
     differentiation that they have to drive for.

Quinta do Lago, 25-26 September 2008                                                                 8
2008 European Press Summit                                                                NetEvents

     And, of course, there is ongoing debate about how will the overall value chain work
     out and distribute. But one thing is for sure, one element of today's operators is and
     stays the delivery of the bandwidth. And they have to optimise this part of their value
     chain in any case.

     Manek Dubash
     Okay. So if we take that on board, you still have all these carriers who are basically
     islands. The carriers cover particular geographical areas in the main. Interconnecting
     those islands, at the moment as I understand it, is pretty much a manual process on a
     case-by-case basis. What can the carriers do to solve this problem because obviously
     this is not ideal from an enterprise perspective? You want to connect the data centre
     in Brazil to one in Manila or wherever. You need some form of interconnect. What's
     happening in that area?

     Bernd Schumacher
     This has a lot of aspects, of course, this question. If you look into multinational
     operators, how they have grown. Very often they came basically from individual
     acquisitions, also in individual countries. So they come from a history where they
     have a huge legacy and a very different legacy. This makes life, of course, not very
     simple. Yes, this is a change in itself. But the common ground is that those legacy
     systems are not ready and are not fit for this challenge. We have ATM around, we
     have frame relay around, we have SDH legacy around. And those systems will not
     deliver the cost efficiencies that are required here.

     Manek Dubash
     But at least they interconnect.

     Bernd Schumacher
     Having said this, interconnection to those legacy systems will be given by, for
     example, [circuit innovation], legacy in the working is a key part. And of course it
     has to be end-to-end tested. But this is all done. And from a technological point of
     view this is solved. If we look for synchronisation part on E1, on T1, this is all
     basically today. And they simply have to overlay more efficient infrastructure and the
     interconnection part to legacy will be given.

     Manek Dubash
     Yes. I agree that the interconnection between the legacy networks is pretty much a
     done deal. But is the interconnection between those islands of Carrier Ethernets
     looking forward.

     Bernd Schumacher
     Yes. On this side, obviously standardisation plays a very important role. And this
     standardisation process is ongoing. Of course we are fully committed here on the
     standardisation progress. We play actually a very active role here in the

Quinta do Lago, 25-26 September 2008                                                             9
2008 European Press Summit                                                                    NetEvents

     standardisation. By the end of the day the complete industry and vendor community
     has to drive towards standards. But it's always like this, the early bird catches the
     worm. And we have to move very fast to serve the immediate needs of the customer.
     And we have to maintain then the balance, of course, to make sure that the complete
     ecosystem moves in a, let's say, synchronised way in this direction. But
     fundamentally I believe and am absolutely convinced that there are no big other
     choices in terms of creating those highly efficient network architectures.

     Manek Dubash
     Okay. So we've talked a lot about where we are and where we're trying to get to. The
     hardest thing for any carrier is how do you do that. How do go from here to there?
     What do you think is the biggest single challenge that the carriers face?

     Bernd Schumacher
     The biggest single challenge I would say is implementation speed. I think it's
     implementation speed to make sure that they can provide those services very fast,
     very effectively into the hotspots in the network to roll out the services really very fast
     and to migrate also services from the legacy world onto the new architecture.

     Manek Dubash
     So speed as in getting ahead of the competition, for that reason?

     Bernd Schumacher
     Yes. One of the real, the biggest fear points today on the operator, it's really churn on
     the customer base. It's a fierce competition. So the one who runs fast will make the

     Manek Dubash
     Okay. I'd like to throw the questions open to the floor. If anyone has any questions,
     please indicate and a microphone will appear magically in front of you. A question at
     the back there.

     Bob Emmerson – The TMC Group
     Yes. Good morning. I think in the Carrier Ethernet, I think the benefits of Carrier
     Ethernet are compelling. And I think there are something like 150 members in the
     Metro Ethernet Forum. I'm Bob Emmerson. I'm a freelance writer. The benefits, as I
     said, are compelling. I think there are at least 150 vendors in this space. Maybe more.
     And Nortel, I think, have just pulled out. Do you think there's going to be a massive
     shake-up in the Carrier Ethernet space in the coming years? There just seem to me to
     be far too many vendors there.

Quinta do Lago, 25-26 September 2008                                                                10
2008 European Press Summit                                                                    NetEvents

     Bernd Schumacher
     It's very hard to speculate here obviously on the future and on whether a massive
     shake-out is going to happen here or not. I think what we do see right now is that
     some technologies or standardisation parts are emerging. And basically the right
     technology choice in terms of what is supported by our overall ecosystem is extremely
     important. And from this point of view there might be some contenders in this race,
     let's say, that will not make it. But I think in terms of overall landscape, I think the
     industry will manage towards a clear standardisation process here. And we will have,
     like we have it today in the transport networks, we will have a lot of significant
     vendors around here.

     Manek Dubash
     Let me just follow up to that. What stage do you think we're at in terms of Carrier
     Ethernet equipment providers? Are we into the consolidation phase or do you think
     there's still lots of room for expansion in terms of number of companies?

     Bernd Schumacher
     I wouldn't speculate here on further consolidation steps right now. I think this is by
     far too early and it's very hard to judge on this right now.

     Dean Bubley – Disruptive Analysis
     Dean Bubley from Disruptive Analysis. Are there limiting factors on mass Carrier
     Ethernet rollout, particularly to homes and individuals? Are they related really to the
     technology or is it more to business and governmental issues in different countries?
     My understanding is that the main cost of, say, putting in fibre to the home is digging
     up the road. Really have we got problems here with legislation and regulation around
     things which are outside the technology domain? Are they the limiting factors?

     Bernd Schumacher
     There are such limiting factors. But again, also here, I would say on a global basis the
     world is quite colourful here. And, of course, nobody in the operator side would like
     to pay for digging up the streets, putting in fibres, investing up front, and later on this
     specific operator has to open this all up to the competition. Of course, there are issues
     around legislation here that have to be solved. But, again, this is not only related, for
     example, fibre access to the home. Also if you look to available broadband
     technologies over copper. If you look to broadband technologies, if you look to LTE
     and the bandwidth uptake comes along such technologies that might have less
     restrictions and, let's say, less influence from a legislation part.
     I would say, on a global basis, yes, those things are around and do exist. It depends
     on the country by country that you look into. But at the end of the day the topic will
     be that new services will have to be provided, that there's a need for those new
     services, that there's a willingness to pay for the services. And here it ties directly in

Quinta do Lago, 25-26 September 2008                                                                11
2008 European Press Summit                                                                    NetEvents

     again into the cost around providing those new services. And this is what Carrier
     Ethernet transport has to optimise and will optimise.

     Manek Dubash
     Bernd, do you think the industry is doing enough to encourage regulators to, for
     example, this was something people were discussing yesterday, gas utilities come
     along and dig up the road but then they go and close the hole again, whereas actually
     what they could have done or somebody could have done was say well okay lay fibre
     while you're there, while there's a hole in the ground. Is there enough going on from
     the perspective of persuading legislators to put regulation in to allow this to happen in
     order to speed the rollout of this?

     Bernd Schumacher
     As said, it is not only in this way that putting fibre into the ground is the only cure. If
     you look to the next generation network, of course, at some point in time you will
     have deeper fibre penetration. This does not necessarily mean that this fibre
     penetration goes directly to the home. It will be closer, for sure, to the subscriber.
     There will be a mix of different technologies around there.
     And, as said, I think we should not oversimplify it only with focusing on deep fibre
     rollout. As said there are different access technologies available already today that
     can enable the bandwidth that we were talking about, from a fixed network as well as
     from a mobile network perspective. And I don't believe, at the end of the day,
     legislation around this will be the hindering factor, because the driving forces are
     simply there. It is the services. It is IPTV services. It is video content. And it is the
     urgent need today on the enterprise sector to interconnect basically in a distributed
     working environment, whether this is data storage, whether this is centralised IT
     applications. So the need is out there already today.
     Take a network like COLT, this is in existence today. The services are running today
     over this network. So this is already a reality. And of course it starts very heavily
     from the enterprise business services part.

     Bryan Betts – Techworld
     Bryan Betts from Techworld. Couple of points. One is still talking about services
     here. And I think, Bernd, you've already said, on the one hand you're talking about
     services, on the other hand you're talking about the carriers needing to become more
     efficient. Why are we still talking about services? They are going to become pipes
     aren't they? They are just going to be pipes in the future.

     Bernd Schumacher
     You might say that they are going to be pipes only. But the need for efficiency on
     those pipes is extremely high. And if you take a specific example of enterprise
     services, you might argue that this is a capacity pipe. But at the end of the day,
     specific service parameters and quality of service parameters are associated to those

Quinta do Lago, 25-26 September 2008                                                                12
2008 European Press Summit                                                                  NetEvents

     pipes. It's about protection mechanism. It's about resiliency. It's about quality of
     service parameters. It's about flexibility of bandwidth provisioning. It's about
     scalability of the bandwidth. So a lot of things come along with this.

     Bryan Betts
     Why are you still betting that carriers imagine that they can pretend to be a service,
     pretend to offer services?

     Bernd Schumacher
     Because they will be a key service provider in this field. And they own the
     infrastructure today. They own the customer. They have the customer access. And I
     fundamentally believe that they have a very good starting point in this direction.

     Tony Savvas – Computer Weekly
     Anthony Savvas, freelance from Computer Weekly. I don't suppose you mind who
     you sell your kit to. But is there more room for government to actually get involved
     in getting things going? In the UK, I think I'm right in saying that the government, or
     one spokesman has discounted any action of actually rolling out fibre. But in other
     countries there are bits and pieces going on.
     I just wonder what you think the industry can do to make sure it can happen, going on
     what Manek was saying. How long are we going to be waiting to get some sort of
     uniformity in the services you're suggesting, like 100 megabits to the home? It's
     happening in little hotspots. But what can the industry do to involve the government,
     get the regulators together, because you have a lot from the European Commission,
     for example, saying that they believe in an equal society, partly through the provision
     of IT. So I'm just wondering what you think the industry can do to get things going.

     Bernd Schumacher
     Yes. If you’re addressing the issue of governmental support and, let's say,
     governmentally driven projects, this is one side of the equation to spark demand, to
     put upfront investment in and to basically to promote such new services infrastructure.
     You could see like the roads and like the highways as a basic infrastructure. And I
     think this understanding is quite widespread, basically within all governments of the
     world. We have done a very interesting study here as well, and it resulted also in a
     call for action in significant papers and studies that we have discussed also intensively
     with government bodies. And I think there's a common consensus that basic
     infrastructure is, at the end of the day, a competing advantage for economies that we
     have to drive, we have to have.
     What can we do as an industry? I think we, as a telecommunication industry, we have
     to work on the efficiency of those solutions and the cost point of those solutions,
     because there's always two angles. There's a kind of inflection point of upfront
     investment that is necessary that might be supported in whatever means from a
     governmental perspective, that's one side. But the other side, I would say, it's deep in

Quinta do Lago, 25-26 September 2008                                                              13
2008 European Press Summit                                                                      NetEvents

     our roots and in our heritage and our mission to make sure that the entry hurdles of
     providing such solutions are lowered. And it has to do with efficiency of the network
     designs and the cost points around it and the flexibility, how to manage those
     networks, how to use them, let's say, multipurpose for different services. I think we
     have to work on both sides.

     Steve Broadhead – Broadband Testing
     Steve Broadhead, Broadband Testing. My feeling here is I started looking at Metro
     Ethernet, as it was called, in around 2002, 2003. And we were talking about services
     then. And I just don't see a huge amount has changed. So my question is, how long
     has Carrier Ethernet got to actually succeed, in the sense that if you look at the LAN
     environment, for example, Ethernet has always succeeded because it's the cheapest,
     dirtiest technology, and ultimately it wins, regardless of ATM, Token Ring, etc.
     However, if you look in the WAN Internet environment, there's an even cheaper,
     dirtier technology called ADSL that has been winning and that continues to increase
     in performance, etc.
     And I'm just asking the question, for once, is Ethernet actually being beaten at its own
     game by an even cheaper, dirtier technology that can be implemented right now and
     everything is happening. How long has Carrier Ethernet got to actually succeed?

     Bernd Schumacher
     I see it a little bit differently than how long does Carrier Ethernet have to succeed. It
     has started already today. And it is starting in significant network rollouts. It is
     starting with Carrier Ethernet transport ingredients in all our products today and in
     real life networks. I think it has started already today.
     And I actually don't see, right now, in terms of technology that is available, any other
     different technology that can cope with this projected bandwidth demand. There is no
     other technology available. You called it Ethernet in the LAN, dirty and cheap. Yes,
     it has to be cheap on the one side as well. But there are different parameters when
     you bring Carrier Ethernet and Ethernet technology into the WAN environment. So
     it's obviously the reliability and it's the quality of service aspect. It's protection aspect.
     It's manageability, end to end, basically point-and-click provisioning.
     So all the aspects that are also stated out of end user perspective. When you want to
     have a service activated, how fast do you get it? How reliable do you get it? What
     are the quality of service mechanisms around it? Do you get guaranteed SLAs? So
     you can't compare it one on one to LAN side. And I do not see any other technology
     available today in the market that can solve exactly those challenges that I have just
     mentioned, in combination with the bandwidth growth. So it has started already today.

     Steve Broadhead
     It’s just the speed at which it has developed. I don't disagree with anything you've
     said there at all. It's completely logical. But it's just the speed at which it hasn't
     developed that worries me, and also the fact that there are more and more optimisation

Quinta do Lago, 25-26 September 2008                                                                  14
2008 European Press Summit                                                                 NetEvents

     products around to make more of what bandwidth is already around. That's one of the
     big growth areas that I test in. And everything has a limited time span in which to
     succeed. And it just worries me that the message, for whatever reason, isn't getting
     across quick enough to the carriers, telcos, service providers, etc.
     And so, just to reiterate what everybody else is saying, there has to be a big push from
     somewhere to put this in people's faces and say look, do it and do it now. But where
     is that going to come from? Who's going to provide that big push? Because all your
     vendors have been trying to do it, but for me it hasn't happened. And I've actively
     been involved in that process of testing and stuff, but it just hasn't happened.

     Bernd Schumacher
     The push has to come, at the end of the day, from having services where there's a
     willingness to pay. This is obviously the key ingredient of it. And a very integral part
     of this push obviously is to get the most efficient network design so that the complete
     value chain, from overall perspective, from anyone's perspective, can be an extremely
     efficient way.
     And if we look to adoption speed, of course that's always something that is, I agree
     with you, hard to predict if you look into the future. But I can say what I've shown to
     you here in terms of predicted bandwidth growth is a quite common sense and shared
     view basically amongst the complete community and all major network operators.
     And we can see it also if you look, for example, on the optical rollout, how capacity is
     being prepared for us in the core networks.
     You can see it on evolution of the technology and how 10G today in the network on
     DWDM, 40G is today a reality in the networks where we have today basically largest
     rollout of 40 gig technology today. You can see with 100 gig that is not already only
     on the horizon, it is already there. And you will see this also with quite some
     interesting press announcement today, that 100 gig is becoming also there a reality of
     real and live networks, not only in a lab space. I think I see quite some strong signals
     that this bandwidth uptake is happening already today.

     Manek Dubash
     On that optimistic note, since there are no further questions, I'd like to thank you,
     Bernd, for an interesting presentation. Thank you.


Quinta do Lago, 25-26 September 2008                                                             15

Shared By: