Nortel: Top Dog, but for How Long?

For the second time in two weeks, Nortel Networks Corp. (NYSE/Toronto: NT) has issued a press release claiming the leading market position in sales of "optical Internet" gear.

But scratching the surface of both announcements reveals that Nortel's boasts, while valid, gloss over its continuing small share of the high-channel-count DWDM market. Further, analysts say Nortel must act vigorously to keep its overall place.

Let's take it from the top: In last week's announcement, dated February 19, Nortel says market research firm Ryan Hankin Kent Inc. (RHK) reported Nortel to have a 61.4 percent share of the North American long-haul DWDM market -- double the share that RHK gave it one year ago.

The overall North American market for DWDM in 2000 was $7.7 billion, according to RHK. The firm divides up the DWDM market into two segments, however, and Nortel doesn't dominate them both.

The first segment in RHK's taxonomy for 2000 devotes $6.1 billion of the overall DWDM market to systems with channel counts lower than 40 -- a segment that Nortel led with a 72 percent market share, according to RHK's research.

Low Channel Count DWDM Systems RHK's second segment includes so-called high-channel-count systems, with more than 40 channels -- a market RHK sizes to have been $1.6 billion in 2000. Nortel has just 10 percent of this chunk, which is led by Ciena Corp. (Nasdaq: CIEN) and NEC Corp. (Nasdaq: NIPNY).

High Channel Count DWDM Systems Nortel's lack of presence in this division could be cause for concern, since that area is slated to be among the fastest growing in optical networking. For example, in publicity for a separate report on optical components, RHK cites "performance improvements -- such as higher channel counts" -- as drivers for triple-digit growth in DWDM systems over the next three years.

It's also worth noting that Nortel's share of the high-channel-count DWDM market didn't increase from 1999 to 2000 (see Nortel Spins Past Ciena).

The conclusion? Nortel must hustle if it means to keep its top spot in DWDM. It could lose out to other players quickly if high-channel-count systems start to predominate.

This analysis is borne out by research touted in this week's market share announcement from Nortel. On Tuesday, the company announced that research from The Dell'Oro Group shows Nortel to be top dog all 'round.

DWDM Long Haul Market DWDM Metro Market Sonet/SDH Market But report author Shin Umeda issues some caveats.

First, Dell'Oro does not use the same type of breakdown in its figures that RHK does. Instead, the firm sticks to sales counts it gets directly from the vendors in making its tallies. In the case of Nortel's long-haul DWDM figures, the vendor gloms together sales of its S/DMS Transport Node, an older product that works only with Nortel's Sonet add/drop multiplexers, and its newer Optera LH 1600, which isn't dependent on those ADMs.

It's an important distinction, some say. Nortel's products based on legacy Sonet and SDH (which Umeda estimates account for over half of all Nortel's optical revenues) are predicted to start phasing out as carriers demand next-generation gear with multivendor compatibility and higher channel counts.

Umeda says Nortel refuses to divulge which percentage of its long-haul DWDM revenues come from which products. Nortel confirms this. But a Nortel spokesperson says, "Optera LH definitely makes up the majority of our revenues in this space, and even without [the older products being included] we'd still have the number one spot."

For the record, Umeda says Dell'Oro has included only revenue from equipment in its reporting; no service and support is included. Also, he asserts that he's not double-counting DWDM equipment; what's counted in long haul does not appear in metro.

Other analysts confirm that Nortel needs to avoid complacency about its market position, in light of the changing nature of the DWDM market. "Nortel has a lot of old equipment that's intimately linked to its old-style ADMs," says Mark Storm, optical networking program leader at research firm Frost and Sullivan. "The [S/DMS] is where the bulk of the revenue is, and they're challenged to make the transition [to next-generation gear]."

He also says market researchers themselves need to revamp their taxonomies. "There's a key transition going on from Sonet/SDH to next-generation gear. It takes a while to slice and dice it all." He says researchers must avoid inflating the numbers by counting too much, while making sure not to miss the key trends by forcing new gear into old categories.

-- Mary Jander, senior editor, Light Reading http://www.lightreading.com

<<   <   Page 4 / 4
Cerent Rules 12/4/2012 | 8:49:38 PM
re: Nortel: Top Dog, but for How Long? JohnT:
We would definitely like to see your results.
Unlike what the Lightreading discussion board writers have to say, your work is not worthless.
What school are you at by the way?
wdog 12/4/2012 | 8:49:37 PM
re: Nortel: Top Dog, but for How Long? ased on the specific comments made by Don Smith, I believe the Nortel web site is wrong regarding the ability of the HDX to groom to STS-1. I have always heard it refered to as the bandwidth manager for their OC-768 family. The DX does groom to STS-1, but is more like a traditional 3/1 cross-connect and only supports a very limited number of OC-48 ports per bay, and I don't believe it supports OC-192. I have never heard of it making the cut in any competitive situation against the CoreDirector
lightreading 12/4/2012 | 8:49:35 PM
re: Nortel: Top Dog, but for How Long? JonT:

This is AWESOME. There is absolutely nothing wrong with some simple scientific analysis. It will be very interesting to see if and how these things correlate. For all you people who think this is useless, don't worry - if it is it will show up in the data. If not, then you will be glad to know. You can't argue with numbers; they are not biased and have no preferences. That's the beauty of it.

Keep up the good work and critical thought JonT. Since Light Reading reporters and editros love to hide behind this mantra, I would think they should have absolutely no problem in providing you with all the data you need. Not hard dollars, but certainly giving you relative advertising revenue percentages by company should be totally accpetable. If it is not, then there is something else going on and something to hide.

Seek the truth.

prefer_to_lurk 12/4/2012 | 8:49:33 PM
re: Nortel: Top Dog, but for How Long? "The DX does groom to STS-1, but is more like a traditional 3/1 cross-connect and only supports a very limited number of OC-48 ports per bay, and I don't believe it supports OC-192. I have never heard of it making the cut in any competitive situation against the CoreDirector"

The Connect DX is totally designed to support OC-192, and has been deployed in large quantities in numerous long haul networks. I don't where your 3/1 cross-connnect misconception comes from.

Is 72 OC-48 ports per bay a very limited number ?


It will compete with the CoreDirector CI on a head-to-head basis, with the Connect HDX matching up very favorably against the CoreDirector.


wdog 12/4/2012 | 8:49:33 PM
re: Nortel: Top Dog, but for How Long? Sorry if my information was incorrect. It sounds like you know quite a bit about the DX and HDX. Do you know of any situation where the DX competed against the CoreDirector and won? I hear about CoreDirector competing against the Sycamore SN 16000 in many situations, but I've never heard of a single situation where the DX was even considered.
prefer_to_lurk 12/4/2012 | 8:49:32 PM
re: Nortel: Top Dog, but for How Long? wdog:

It's actually not my area of expertise. However, to your question, the DX doesn't (and isn't intended to) compete with the CoreDirector -- so that's why you haven't heard about such comparisons.

The Connect HDX (coming out later this year) will compete head-to-head with Ciena, and we'll just have to wait and see how often it comes out on top.

alloptical2000 12/4/2012 | 8:49:31 PM
re: Nortel: Top Dog, but for How Long?
Most of the facts are well covered about tha capabilities of the DX.

and also, as you said, DX competes head-to-head with CoreDirector CI, is perfectly inline. as far as i know CoreDirector CI is a scaled down version of the Coredirector with half as much(approx 320Gbps) capacity for switching.

According to Nortel's website DX can do 320 Gbps of switching. The port count should be up-to 128 X 128 (max) at OC-48 level.

DX has functionality of cross-connection and grooming at STS-1 level and also traditional transport of SONET in the long-haul networks.
But i am not sure if CoreDirector CI has transport soultion. It is a pure cross-connect. correct me if i am wrong.They do compete only when configured as a cross-connect.

One of the reasons why DX might have not fared well in the competiton was probably due to a smaller switching capability of 160G at that time.

It is only recently that they have proposed the soultion of 320Gbps of switching. They might have compromised on the tranport piece of the functionality to achieve the density and the switchining capabilities at 128 X 128 OC-48.
I am not even sure if 320Gbps soulution is available yet. the 160G soulution is available for sure.

and the HDX would be Nortel's best bet aganist CoreDirector. But, it is going to be a while before we see that.

I see the following two being launched and competing head-to-head with CoreDirector towards the middle and end of the year:
1. Nortels' HDX
2. Tellabs's 6500/6700 (tellabs has been working on this for while now.)
3. Sycamore's 16000

any comments and feedback welcome
jppr17 12/4/2012 | 8:49:30 PM
re: Nortel: Top Dog, but for How Long? I believe segmenting the DWDM market only in terms of number of wavelengths is not precise.

The important value is the combination of both total transport capacity (per channel TDM x # channels in terabits per second) and distance reach (without EO regenneration) that a product can deliver.
y2k 12/4/2012 | 8:49:27 PM
re: Nortel: Top Dog, but for How Long? "2. Tellabs's 6500/6700 (tellabs has been working on this for while now.)"

I understand that Tellabs' 6000 series can have up to 3,000 ports of OC-48 and the granularity is at OC-48, similar to Cisco/Monterey and Tellium.

Tellabs is late but not to be ignored. They have a great installed base with RBOC's and they know how to sell to them. In any case, they do not compete directly with CoreDirector nor DX/HDX except when these products are used strictly for mesh restoration and protection.

alloptical2000 12/4/2012 | 8:49:17 PM
re: Nortel: Top Dog, but for How Long? Hi,
I am not really sure if that is the case.
Following is the link to the telllabs 6500.


it does list all the calpabilities of CoreDirector. especially the STS-1 based grooming and switching.

<<   <   Page 4 / 4
Sign In