Article Info

Everyone Wants an NGN

Those of you in the component, equipment, or carrier businesses should listen to the latest strategy presentation from Carphone Warehouse, a competitive carrier that offers broadband, voice, and mobile services in the UK. It offers a stunning picture of how low the economic barriers to entry have become when building a greenfield network. Now that the price of owning one has dropped, everyone wants a Next Generation Network (NGN), and no one wants an albatross of old telecom kit.

Amidst the Herculean efforts of British Telecom to roll out a nationwide forklift upgrade of their network (see BT’s 21CN – Reversing a Victorian Tradition), CPW is building their own mini-21CN and plans to bypass the use of BT’s network for transport services. This would make Carphone a completely vertically integrated carrier from a retail store selling services all the way to big-iron DSLAMs sitting in a local telephone exchange. It’s an extraordinary endeavor.

I suggest everyone spend 25 minutes and listen to the Carphone webinar overview of their NGN. (note- you need to skip to the section on the network)

One very interesting slide:

How can they achieve a network of similar function with substantially less cost? They are not burdened by 30 years of legacy protocol and equipment and can start fresh.

Carphone cites the rapid drops in equipment costs as a key enabler, as well as the lower costs of owning and operating your own dark fiber. They also cite that rapidly growing traffic requirements make outsourcing the connection un-economical.

This decision by Carphone is an excellent example of how the telecom capex environment is changing. Everyone is aware of the fundamental benefits of new equipment and protocols but they believe carriers cannot rapidly change the structure of their infrastructure. As a result, people (including myself) refer to telecom equipment revenue streams as annuities that can be milked for long periods of time.

I’m beginning to believe this assumption no longer applies. The reduced cost of owning and operating new equipment is so much lower than the status quo that carriers can no longer justify not capping expenditures on legacy equipment.

Verizon FiOS, Enterprises leasing and operating their own dark fiber, Cable VoIP, and IP-PBX’s are all examples of where a radical departure from the status quo delivers cost savings that are too compelling to ignore.

Equipment and Component vendors that do not have both feet in businesses that are built around NGN architectures will face faster declines than they expect. If I had a a legacy telecom ‘annuity’ business I think the time has come to take a lump sum payment.

What about the nightmare scenario: Have hardware providers become so good at delivering value that the total market for telecom hardware will decline even as demand increases?

I remain optimistic that networking equipment is no different than computing equipment, and that driving costs lower results in an even greater expansion of unit demand. That being said, as legacy equipment is thrown overboard, we may experience a one-time discontinuity that will be exceptionally painfully for some vendors.

Full copy of the presentation is here. But the 25min webinar is worth your time.


Comments are disallowed for this post.

  1. This post seems to refute completely your assertion that Sonus is VoIP 1.0 as postulated in the post, “BT’s 21CN – Reversing a Victorian Tradition”.

    Posted by gnevik | May 2, 2007, 9:23 PM
  2. No, it does not. Sonus makes a good softswitch. But they are not as well placed to enable the next set of features that will make consumers gladly pitch their landlines.

    I will agree the article makes a good case for Sonus vs. Ericsson and everything I have heard about 21CN points to Ericsson as being the caboose vendor in the project.

    The flip side of this is Ericsson is working like hell to solve a problem and learning along the way. 21CN is a project that is ‘too big to fail’ and Ericsson will emerge stronger. Ericsson has a bigger installed base to migrate than Sonus, something they can monetize in a way Sonus cannot.

    I can’t believe how devoted the Sonus internet message board jihadists are. They are third behind Apple (#1) and Akamai (#2). You wouldn’t believe the hate mail I got. I AM NOT NEGATIVE ON SONUS.

    Posted by Andrew Schmitt | May 3, 2007, 7:42 AM
  3. Andrew,

    Your opinion is shared by many in the industry, Sonus investors just disagree with it. This is not only common for a dispruptive little company, it is standard practice. You can look back to MSFT as PC OS’s developed, QCOM in wireless, you name it, there seemed no way incumbents could be un-seated. Sometimes(maybe often) incumbency does prevail, it is a powerful force to overcome.

    Sonus has won significant core network upgrades against well ingrained incumbents – KDDI and Cingular are the wireless highlights, VZ and DT are expanding networks with Sonus in the wired arena. Access wins at incumbents have been tough.

    Where I would take issue with your article is you chose not to mention Sonus, though you did display a slide with Sonus listed. After listening to the network presentation from Carphone, I heard multiple references to a simple concept – Sonus replicates Class5 services and Ericsson does not. Sonus scales well, we have no clue if Ericsson scales because the service has not moved beyond trials. The impression the presentation gives and the one you give are not the same. You highlight incumbent operator vs. upstart, we feel it is incumbent VENDOR vs. upstart.

    Of course we are defensive, we see time and time again an incumbent win access business on a promise against our proven solution, any time press or blogs or analysts further that idea we try to get our side of the story out. Here we have a real world example of what happens when one operator chooses Sonus and another chooses a promise from a Powerpoint slide. We’d like to see that highlighted. We feel it is the same thing at Verizon(Nortel) or SBC(Lucent), mounds of press about the contract, barely a peep when the service offering is months or years late to market.

    I will say the Jedi Mind Trick article was great, although it did not reference us directly, I think it represents a mindset at incumbent operators.


    Posted by Cooters | May 3, 2007, 12:24 PM
  4. What I write here is my own opinion, I don’t plan to be a fair and balanced outlet for news (though such a thing cannot exist).

    Reading your comment reminds me that one should never become so emotionally tied to a stock that you ignore the reality of the facts. If carriers are siding against Sonus, as you say, what is the likelihood of a massive decision reversal?

    What if softswitches are obsoleted by the time carriers figure out how to make them work? Why not just put a SIP stack on every client and push the brains of the network to the edge?

    Posted by Andrew Schmitt | May 3, 2007, 3:11 PM
  5. I think that the fact of “carriers siding against Sonus” is something that was in the past. The carriers gave the initial tries to NT, ALU, ERIC … without any real results. And now they seem to finally be seeing that Sonus is the only one that can make it work. Or I should correct again, Sonus is the only company that HAS made it work.

    Your slide shows it all, Carphone began deployments with Sonus AFTER BT started theirs with ERIC and right now BT has 1 exchange in trial and Carphone has 1,100 in service. What does that say?

    I think allot of this has been caught up with SONS vs. APKT. At least for the immediate future I don’t think it is one vs the other. Thinking APKT has success with SBCs in the next few years doesn’t mean that SONS won’t have success with the NBS in that same time frame and beyond.

    And when you say “What if softswitches are obsoleted by the time carriers figure out how to make them work?”. I think the more correct phrasing would be “by the time the INCUMBENT VENDORS figure out how to make them work”.

    The fact is Sonus is the only vendor worlwide to have successful, major, scalable, non-TDM hybrid, carrier class VOIP deployments … in the world. The comparison of BT to Carphone shows this example quite well. Sonus has multiple examples to show of … can you give ANY example of such from ANY other vender? Even one?

    jmho of course. But I do believe that just as you think Cooters investment in SONS is skewing his view of APKT, I believe your bias towards APKT is skewing your view of Sonus, perhaps significantly.

    Again, but that’s jmho.

    Posted by cdt2001 | May 3, 2007, 3:44 PM
  6. I think allot of this has been caught up with SONS vs. APKT. At least for the immediate future I don’t think it is one vs the other. Thinking APKT has success with SBCs in the next few years doesn’t mean that SONS won’t have success with the NBS in that same time frame and beyond.

    I totally agree. They are complementary which is why Sonus resells Acme Packet.

    You are right about my bias. Trust me, I wish I was a SONS investor as of late. If you think the market has a hard time understanding Sonus, try Acme Packet on for size. We should both just give in to the herd and buy Riverbed…

    Posted by Andrew Schmitt | May 3, 2007, 3:58 PM
  7. Actually … personally I think the market can understand APKT more. They have the majority (50% ) of their current market where they sell their products. That is something the street understands.

    The street would much rather see that then see a situation like Sonus where the vast majority, even industry professionals have a hard time seeing how instrumental Sonus is really becoming. As NT, ALU and to this point ERIC aren’t able to come up with real deployments like Sonus has done … they appear to have a good shot at quietly becoming the dominant company in carrier VOIP.

    But when research reports that include Hybrid VOIP deployments in the same catagory as deployments SONS does … all the street sees is Sonus as the laggard. And until it becomes more evident of Sonus’ sucess, and more importantly that sucess is translating into revenue (which b/c of Sonus’ rev recognition practices can actually be lagging 3-4 Qs of actual deployments) … the street still won’t understand Sonus.

    Posted by cdt2001 | May 3, 2007, 4:16 PM
  8. The flip side of this is Ericsson is working like hell to solve a problem and learning along the way. 21CN is a project that is ‘too big to fail’ and Ericsson will emerge stronger. Ericsson has a bigger installed base to migrate than Sonus, something they can monetize in a way Sonus cannot.

    Good software cannot be written by working like hell… you must not be a s/w guy like I am. It introduces bugs and that’s what NT, LU, had been facing for years (and Ericsson is going to face this reality someday, too!).

    You must not be a s/w person or should know how many LOC (lines of code) of softswitch has. It is obvious SONS started when rest were sleeping (LU did try but shutdown R&D as part of their cost-cutting in late 90s). There are millions lines of code in a softswitch; debugging requires execution of this code over a long period of time and testing all functional components. You cannot do replicate all possible scenarios in labs regardless of how big your team is or how smart the heads are. SONS is carrying billions of minutes every month. It has already filtered out all the bugs – you must accept this fact. Rest has not yet gone to deployment let alone facing the bugs and filtering.

    It is a time-consuming thing and has a long development cycle. I wish Ericsson good luck working like hell – a perfect setup to failure IMHO.

    Posted by PakBrain | May 3, 2007, 4:35 PM
  9. Andrew,

    Certainly did not convey an “Incumbent Carriers vs. Sonus” environment, it is “Incumbent Vendors vs. Sonus”, with the carriers going along for the ride off the cliff. The carriers are stuck in the same mindset you are, let’s give this business to unproven or in many cases “yet to be written” solutions and they can “work like hell” and get these services up and running. So in access cases where TDM solutions are currently servicing customers, they just wait and wait. When access service offerings by new operators like Carphone or core upgrades(VZ, KDDI, Cingular) require equipment to function at or above current service levels now, Sonus wins. There is plenty of core business just from the world’s wireless operators to make Sonus a great investment, I’m not stressed about what is going on at BT, VZ, and ATT, I just cannot understand their acceptance of failure.

    The incumbents can’t even get this right and you are already looking to a future where the intelligence Sonus provides moves to the device. Peer into the future too far, you are. :)

    I enjoy the vast majority of your blogging, just think you have this one wrong.


    Posted by Cooters | May 4, 2007, 7:58 AM
  10. I think incumbent operators have no choice as they are hobbled by an installed base. Someone like Carphone has the benefit of starting fresh and this favors new vendors. That was the point of my article.

    While I have everyones attention, a common opinion is that SBC’s have no future and will be obsoleted. Agree or Disagree?

    Posted by Andrew Schmitt | May 4, 2007, 10:49 AM
  11. Agree, only the timeframe remains in question, IMO.

    Posted by Cooters | May 4, 2007, 11:16 AM
  12. I actually disagree. I do think the SBCs have a future … however I do think their time will be limited. I think in the next few yrs their useage will peak out, but I still think it may be a while before they are obsolete.

    But I do again think that the Sonus route with the NBS is where things will migrate, and while both that type of architecure as well as SBCs will have allot of growth over the next few years … as the SBC growth tapers off the Sonus route will still be growing for the next decade or more.

    It just depends how for into the “future” one wants to look.

    Posted by cdt2001 | May 4, 2007, 12:57 PM