This Tues Cisco announced their new addition to the CRS family of core routers – the CRS-3 – billed as a 322 Terabit Router. This means it has over three times the capacity of the CRS-1 which has a maximum system capacity of 92 Tbps. A large percentage of the announcement was spent convincing us that the world needs a 322Tbps router – drawing on data from Cisco’s Visual Networking Index. Nobody’s arguing with Cisco on this point. Mobile data, video content, web-attached toaster, blah, blah blah – we get it. However, just as nobody is actually going to transmit the entire contents of the Library of Congress over the Internet in four minutes (although we will all sleep better knowing that they could), no one is going to string together 72 23 inch chassis, each three feet deep and over six feet high, to achieve the parlor trick of the 322 Terabit router. I’m not saying that we’ll never need a 322Tbps core router. I’m just saying that when we do – it’s not going to be delivered by the CRS-3. Let’s look for a moment at the CRS-1 – the 92Tbps router. Introduced in 2004, the CRS-1 also requires 72 racks to build out to its full 92Tbps capacity. However, while Cisco is keeping tight-lipped on this subject, Yankee Group believes that the vast majority of CRS-1s to be 1-5 chassis systems and we know of no systems over 10 chassis. By the time Cisco customers moved into the double digits of chassis – Cisco introduced the CRS-3 which could deliver three times the speed per chassis. In 3-5 years when CRS-3 implementations are, at most, in the low double digits – Cisco will introduce the CRS-5 or 7 or whatever catchy name the company comes up with. That leaves us with the existential question of: Is a router really 322Tbps if nobody ever implements the full configuration? It does, certainly, allow Cisco to boast the highest capacity core router. However, the Juniper T1600 core router still achieves in half a rack what the Cisco CRS-3 achieves in a full rack. The CRS-3 will be commercially available later this year. Next year, we expect Juniper to introduce a 250 gigabit slot. Yankee Group expects Cisco to continue to lead in overall system capacity – Juniper in per chassis capacity, for the foreseeable future. Read the rest of this entry »
The second interview I managed to do in Lisbon was of Sue White, who is in charge of Alcatel Lucent’s recently unveiled HLN strategy. I asked Sue to explain what HLN was and how it affected Next Generation Access deployment strategies.
I apologise for the intense shakiness on this video. Definitely need to get a zi8!
At long last, March 9 is here. Cisco had a running timer on its Web site counting down to 11 a.m. eastern this morning when they would make an announcement that, they claim, would change the Internet forever.
Based on the hype that Cisco and much of the media created, I’m not sure what I was expecting but my expectations were high. So what was it? Today Cisco launched its new carrier core router, CRS-3. The CRS-3 is the evolution of a product Cisco launched a few years back, CRS-1, which, at the time, set the high water mark for carrier routers. Upon its release, many people chuckled at the concept of a 92 Tbps router, thinking we’ll never need that kind of bandwidth, but what we found was that indeed we do! Cisco has shipped almost 5000 CRS-1’s–clearly, there’s demand.
Ever since the Google FTTH announcement a couple of weeks ago, the internet (and indeed the real world, for a change) has been abuzz with rumours, analysis, speculation and downright madness in some instances. In the meantime, I have written and published my own analysis of the announcement, accessible to Yankee Group customers under the title Google’s FTTH Experiment Could Profoundly Reshape US Wireline Landscape.
This has led me to also take a deeper look at how bloggers, newspapers and tweeters have dealt with this piece of news and I was surprised and – to a certain extent – dismayed as well to see how polarizing Google’s image is. I guess it’s not surprising that a company so successful should generate such strong passions for or against it but I guess I never realised quite how profound the adulation or the hatred could be.
Here are a few of the arguments, rants, points, or just wild guesses that I’ve come accross in the last few weeks on this topic:
- Google’s FTTH experiment will mean nothing because they will pick a place where it’s easy, and that’s not the situation in most of the US.
- Google is after broadband stimulus money.
- Google will purchase on the cheap from Chinese vendors and the pressure on US Telcos as a consequence will push them in the arms of these self-same Chinese vendors, which will in turn kill the US vendors.
- In order for Google to choose your town you just have to change your town name (and that, IMO has got to be the stupidest thing I’ve seen politicians do in a looooooong time…)
- Google will dissapoint so many towns by not picking them that they’ll end up with a PR nightmare on their hands.
- Google’s plan is a conspiracy to kill incumbents by dictating policy goals of Gb/s bandwidth to local server farms that will be (or already are) owned by Google.
Now some of these might be reasonable assumptions, others verge on crackpot theories. What I find interesting is that I saw very little commentary focus on what was actually in the announcement which, until proven otherwise, is still the most likely course of action for Google.
But perhaps more importantly, this exercice of scanning what is being said on this made me realise that Google has a PR problem on their hands. While there is still a lot of goodwill towards the brand out there, it seems to me that there’s also the seeds of a Microsoft syndrom. An increasing number of people think (and say) that Google is after nothing else than world domination and the Orwellian state. Once that starts, it only takes so long for people to start repeating it.
One thing I heard in nearly every discussion, for example, is that Google as your ISP is a really bad idea, they would know everything about your online self. That’s actually sensible, which is probably why Google clearly stated they had no intention of being the ISP. Strangely enough, many people seemed to disregard that particular aspect of the announcement…
What this really highlights to me is that Google has some serious PR work to do. There’s been a history of disregard for people’s concern around personal information, and while Google has (I think) been relatively careful in the way it’s been using that data so far (whereas the late Phorm hadn’t) the fact that Google has or could have access to usage data is enough to make a lot of people freak out.
I wouldn’t want to be in that place, and it certainly suggests to me that if their FTTH Experiment is to be successful, radical transparency might be necessary. But is that something Google can operate with ?
I did a few interviews at the FTTH Council Europe Annual Conference in Lisbon before my zi6 camera failed me. I also think I need to upgrade to the zi8 for stability and sounds reasons, but that’s another topic.
Anyway, here is the first of the two interviews I could salvage, and it’s with Wolfgang Fischer, the FTTH guru of Cisco in Europe:
Internet video, iPhones, explosive growth of mobile phones in Asia, flat-rate broadband pricing… these and more have sent capacity demands on the Anywhere Network through the roof in the past year. Many of the packets these activities generate end up queuing for intercontinental transport via one or more of the Earth’s submarine cabling systems.
In London last week I had a chance to catch up on the implications of demand and approaches to submarine cabling finance from an Anywhere industry insider, Vinod Kumar. Vinod is President and COO of Tata Communications and a long-time communications sector leader. Among many other submarine cable activities, Tata Communications operates SEACOM, the big cable that recently reached the shores of East Africa from Mumbai, opening up network capacity in Africa in a big way.
Bandwidth demand is booming around the world right now. Do we need more undersea cabling?
“If you wanted to write a check for more right now, I’d say the Atlantic Ocean needs another cable. Not necessarily because of bandwidth demand in total, but rather because of the rise in demand at various landing spots. I’d run one from south Florida at one end, to southern Europe at the other. South America needs another, so I’d run a branch cable down there off of the new one.”
Public markets aren’t enthusiastic about financing big speculative projects right now – and the debt that supported private equity backers is harder to get now, too. How are these projects getting funded these days?
“They can take $250M to $1B at a go. In the old days, the way it used to be funded was through the formation of massive industry consortia. Tata [via the 2006 acquisition at its core, Indian state-run long-distance network firm VSNL] was involved in quite a few. You’d get 60 to 80 firms to commit up front to the commercial demand for the capacity when the cable got laid, in order to get the project financed. But these cable consortia are tremendously complicated to manage; for one thing, you need to control how a consortium’s members push for capacity upgrades before the bulk of the project cost has been recovered.”
“Then you had the speculator model, which boomed in the late ‘90s… for instance, the private equity firm Blackstone funding SEACOM. Rather than go through all the hassle of securing demand in advance, proponents of this model had a ‘build it and they will come’ approach. That boom, though, led to several busts, when the investors didn’t sew up enough commercial commitments before proceeding.”
“Tata then pioneered a model which seems to bring some attributes of each of those models towards the middle. Maybe you’d call it the ‘private club’ model. We own the main intercontinental cable we lay, and various landing parties own the various cables that branch off regionally to local waters, like to Vietnam or the Philippines. It’s less complicated than the consortium approach because there are fewer members. The main asset is 100% owned by Tata, but about 60% of the demand for its capacity is covered by the club members who run branches off it. It’s better for Tata, since we ourselves only have to risk 40% of the investment cost and it’s easier to manage a much smaller group. It’s better than the completely speculative model for the club members, since they get to buy the capacity at our cost plus a limited markup, less than 10%, and since Tata is an experienced undersea cable operator, they can hold us to extremely strict SLAs [service level agreements] to get comfort about reliability. Since we’ve started doing that, others have mimicked the model and it’s become pretty popular.”
But with exploding demand, and Tata’s balance sheet, aren’t you tempted to go the speculative route yourselves?
“Traffic is growing at 60% a year — but no one foresaw the irrational pricing that’s driving some of that. We’ll never take a speculative undersea cable project to Tata’s board — because we don’t need to. We tell the board what our own organic load will be, and we can find the rest of the funds to keep it prudent. It’s a small industry. We have investments in almost 80 cable consortia, so people know us, and we’re good at the work itself, like figuring out how to put cables where other cables aren’t. Those sound like small details, but the earthquake off Taiwan earlier this week disrupted several cables – not ours.”
My conversation with Vinod was on a day when he and other members of the firm’s management team were showing how far the formerly India-only operator has come in the provision of global connectivity. Vinod’s ambition: for Tata Communications to become ‘the Singapore Airlines of network services’. In an episode that reminded me of those times when you suddenly start hearing about the same movie or restaurant multiple times within a few days, a lot of the rest of our talk was about the new models for wholesale network services. I’ll do another post on that shortly.
I’m finally back home after over a week of traveling in Europe. Bloggers Pauline Rigby and Carlos Bock, amongst others, have already expressed their views of the FTTH Council Europe Annual Conference in Lisbon, so I’ll try not to overlap too much of what they covered, but will nonetheless share my own thoughts.
I don’t fully agree with Carlos that it’s been a year lost, but I think he has a point when he says that facially there was very little different to be heard in Lisbon this year compared to Copenhagen last year. I’ve always said and written that FTTH deployment would be a slow affair and we are certainly seeing that now. The fact that the numbers have hardly budged in Europe is certainly depressing, but it doesn’t mean that nothing if shifting in the background.
Where I do agree with him is that I sometimes feel frustrated with a certain lack of vision from many vendors who form the bulk of this show. When I say lack of vision, it’s really about vendors looking beyond their concrete interests in selling gear and understanding how this market might happen. I addressed some of these issues in my keynote speech on the Benefit Compendium – a study that I have undertaken for the Council alongside Roland of Idate – but essentially, with the exception of Alcatel Lucent and Cisco, I’m not seeing anyone tackling the issue yet. I’ll have video-interviews of executives from both companies up later this week discussing these issues.
I had a lot of interesting discussions behind the scenes though which suggest that there is some awareness that the way the market is being developed needs to evolve. The reluctance of most incumbents to move is becoming clearer every day, and the lack of attractive services is also increasingly glaring. That suggests two topics that will be key in the next 24 months: policy and services.
As I’m fond of saying and did indeed say at a DSM Desotech dinner that I was invited to give a speech at
“Deploying the infrastructure is actually the easy part: it can be solved just by throwing money at it. Getting customers to embrace the new infrastructure is the real challenge.”
On the policy side, it’s becoming increasingly evident to me that the incentives put in place by regulators, governments and the European Commission to kickstart FTTH deployment in Europe have largely failed. FTTH is only emerging in countries where the market dynamics make it imperative for the incumbent to respond to competitive NGA initiatives, and even then only as fast as the threat really is. In every country in Europe where this is not true, deployment is at a standstill. It seems to me that the powers that be need to acknowledge this and look at easing local government involvement one more step to create that impetus.
On the services side, there was a sign of hope in Lisbon, and that was due to Google’s FTTH announcement the previous week. Irrespective of the aims of Google on shaping the North American policy agenda, the fact that they want to experiment with fiber grade services is a positive sign. It suggests that they consider the aggregate number of fiber users worldwide will reach critical mass in two years and be worth addressing as a specific market. That should in turn spur the telcos on to accelerate service deployment for their own FTTH networks when they have them…
Graph of Car Locator Sales
Eddie Kim, author of the Android Car Locator app, just announced today that his app has reached a new milestone of $13,000 a month in revenue. Now that may be chicken feed next to what top performing iPhone apps do (the record-breaking PopCap game Plants versus Zombies for iPhone, for example, just went from zero to $1 million in revenue in 9 days). However, Eddie makes a more important point farther down his post:
Read the rest of this entry »
On the 24th of August 2007 an earthquake that measured 8.0 on the Richter scale hit the coast of Peru. The epicenter was just off the coast and located about 150 kilometers SSE of the capital city Lima. Over 150 people died. There was some damage in the capital but most was confined to the towns on the coast.
The mobile networks collapsed due to damaged network infrastructure from the quake, the loss of electrical power and the sheer volume of calls to the affected areas from concerned friends and relatives.
Even while the operators were trying to put their networks back together the Peruvian government started a public firestorm over the outages. “This should not have happened!” was the general tone. The President, Alan Garcia ordered the regulatory bodies to investigate and multi-million dollar fines were bandied about. Operator management was distracted by the public savaging and the need to respond to government officials at a time when they should have been concentrating on restoring services.
Seven months later, the regulator fined Telefonica, America Movil and Nextel a collective 2.8M Peruvian soles or just over USD300K each, an amount that maybe looked good in the newspapers but was nothing more than a face-saving slap on the wrist.
Last Saturday morning (Feb 27 2010) an earthquake measuring 8.8 hit the coast of Chile. Since the Richter scale is logarithmic, the Chilean quake was 6 times as strong, heavily damaging parts of the capital city of Santiago some 317 kilometers to the north of the epicenter. As of this morning (Mar 1, 2010), over 700 people have lost their lives.
Again the mobile networks have collapsed. Again for the same combination of reasons largely beyond the operators’ control. The most affected parts lost internet service but this was restored fairly quickly to Santiago and social networks have pumped out amateur video of the damage.
Since Sunday morning, the country’s chief regulator – Subtel’s Pablo Bello – has been answering customer complaints via Twitter, patiently explaining the reasons for the problems and passing on useful messages from other citizens. The chief problem as he reports it is the lack of electrical power to sites so he has been working to getting vital resources directed to the operators. He has not yet breathed a suggestion that the operators could have done anything in the face of a natural disaster of this magnitude.
His followers on Twitter have received constant updates on his activities to support the restoration of the network and he appealed on the Internet for more satellite telephones. Telefónica has responded. Today we got a region-by-region breakdown of the damage including estimated percentages of the number of base stations affected. Nothing keeps the public from panicking like a constant stream of information from its leadership.
On March 11th there will be a change of government in Chile and Pablo will most likely be out of a job. The new president is center-right and Pablo is closely associated with the outgoing center-left coalition.
I don’t agree with much of what Subtel decides. It is far too quick to intervene in competitive markets for my taste.
But this kind of leadership in a crisis and clear-headed understanding of the issues demonstrates a level of executive skill that is often lacking in many businesses.
Pablo Bello deserves to get a good job somewhere when all this is over. It will be a loss to the Latin American industry if it isn’t in telecom.
Over a fun breakfast last week, I chatted with John Bruggeman, CMO of Cadence, the electronic design automation firm. Just back from Mobile World Congress in Barcelona, I was talking about the battles in the mobile revolution.
John says there are three significant battles still underway in that sector that have do-or-die stakes for the businesses in the actual battle: the mobile operating system (Nokia, Google, and Microsoft — the latter making another run at it with a rethought Windows Mobile), the mobile device platform (the usual handset suspects, Apple, and possibly some daring consumer electronics players), and the prevailing semiconductor architecture — which he sees as boiled down to a question of whether high-performance Intel processors make inroads against the widely used low-power ARM architecture.
Who’ll win that third battle, I asked. The ARM platform has a massive lead in the mobile space, its core IP going into the processor for virtually every handset sold in the world. “Intel is smart, has loads of cash, and knows this is a long-term game,” said John. “Over the next 5 years, they will co-exist. Beyond that, these two worlds — low-power handsets and high-performance portable computing — bleed together. The devices following that time period won’t be about fast web page refreshes; they’ll be about transactions, making fast hits on cloud-based data. When that happens, our mobile devices will want both low power and performance.” Given the time parameters involved, he doesn’t count out Intel’s push to take its desktop/laptop dominance into the smaller more diffused computing domain.
From that topic we wandered over to one that’s a new favorite of mine: that 2010 is the year that mobility as a business issue rises to the boardroom. My logic goes like this:
- The commercialization of the Internet first hit businesses as an external, largely superficial change, in which they essentially stapled websites to their existing operations.
- But the subsequent maturation of Internet computing compelled those same businesses to pull the net throughout their activities, affecting supply chains, marketing and sales, manufacturing, and virtually every other function in the company.
- The mobile revolution has begun similarly. Most major enterprises at this stage have now begun to create mobile experiences for their customers (although, as Carl Howe’s reports on mobile web experiences establish, at widely varying levels of quality)
- The diffusion of the impact of mobility will be no different than that of the Internet. Thus, corporate board members should begin considering how strategically their enterprises’ leadership is thinking about mobility. How else will governance insure that the business is pushing the leverage of connectivity into every nook and cranny of its operations?
Today's version of a mobility officer
John bought it — and he took the thinking a couple of steps further: ”The first automation of business in the 20th century happened with the advent of mainframe computing. The central information systems function arose then. The re-automation of business, driven by desktop computing, pushed IT further out into the business and, organizationally, led to the rise of the CIO. What you’re talking about — the rise of mobility as a strategic issue for businesses — will mean that we’ll see the rise of a Chief Mobility Officer.”
Fascinating idea, and one Yankee Group will pursue in a research report over the next few months with Josh Holbrook taking the lead. But beware: John followed his prediction of the emergence a new type of corporate CMO with this one: “Sadly, many businesses who take this step will put a networking guy in the job. What they’ll need will be an imaginative business person, someone who’s able to look at all the activities of the business and re-think them completely.”