While our fund has done very well with Ascend (In since february at $24) and has gotten burned with NN (rode it down from $47 to $30 last fall - winter) we try not to have favorites, and I would love to jump in to NN. We are a little gun shy and will wait until Mr. Lutz shows more concrete progress against his cost improvement goals.
I'm glad you explained your bias. This makes some of your comments far more understandable.
Nevertheless, Tam Dell'oro has a spotless reputation for digging deep with impartiality. Our experience has been terrific - I never find obvious errors in her data, as I have with IDC, DataQuest and others. In time, earnings will show which source to be more accurate, I have my bias, you have yours.
I'm sure Tam is as honest as anyone can be. However, I also recognize it's close to impossible to get projections from telcos and equally difficult to get honest numbers from vendors. This has nothing to do with Tam or any other researcher, it's simply a fact within the research industry.
As for whether it's more significant to get contracts renewed than to win new contracts, the MO of this industry is that most of the time incumbents win follow-on contracts - even when they are percieved as lagging behind on technology (e.g. Cisco frame relay at T). . . .
This seems to be positive. For example, 90% of NN's ATM sales are to their TDM installed base. Hard to argue with a captive audience.
Ascend has also been taking care of business in installed base - BellAtlantic (including out of territory), BellSouth, GTE, WorldCom. NN has done almost as good a job, but lost Frontier, LCI and USWest.
I understand USWest has not been lost. NN is still doing business there. They also recently won BellSouth International.
The ATM/frame relay market will grow ~30% next year. The question will be who will increase market share, who will decrease or stay the same. With the new contract wins combined with maintaining business with an impressive list of incumbent customers, Ascend is positioned to gain significant market share. Even Vertical Systems Group shows that Ascend has gained more market share than anyone else between 97 and 98. Given that the contracts they signed deploy over 12-18 months with significant add-on business virtually assured over the next 4-5 years, the momentum of Ascend's market share improvement will persist. I believe NN will have fairly stable market share over the same period based on its book of contracts won - this still means its business will grow ~30% with likely improvements in cost structure.
What you say of Ascend is equally true of NN. And as for momentum of market share improvement, I wouldn't bet the farm on Ascend without first studying the master plans of the tier-one vendors.
As to whether these customers are large or fast growing I would say both. Ascend has a mix of smaller, fast growing customers (Qwest, L3, Williams, Frontier) and big, established incumbents (BellAtlantic, Bell South, GTE, MCI/WorldCom, AT&T, NTT). Given that North America sales of ATM continues to outgrow Europe, Asia and ROW, right now I would much rather have Ascend's customer base than NN's ( and NN's over anyone else BTW)
I don't have the expertise to analyse the customer bases on a global scale. I played ASND from 22 to 49 and got out b/c I thought they were ahead of themselves. As for NN, the playing field is shifting and based on conversations with analysts and reports from the company's recent road show to Boston, I think I've made a wise choice.
I have talked to most of the North American customers for both these excellent companies. While as you say, NN customers are pleased - many Ascend customers are absolutely ecstatic. Talk to Williams or Frontier. Try to find a core network engineer at T. Terry Matthews likes to talk big about how there is no competition with NN network managment, but that is not the way customers feel about it. I asked the customers - NN had a big advantage 2 years ago, the customer perception is that Ascend's NavisCore has bridged the gap.
You must keep busy --- talking to all their customers? How do you do it? And how do you judge between "pleased" and "ecstatic?" Terry's never suggested NN didn't have competition but he does maintain no one has a complete multi-service solution that's as robust.
NN has announced it will integrate Cambrian gear, but the biggest part of the value-add is to eliminate expensive SONET equipment. Whether the DWDM circuitry is on shelves inside the ATM cabinet or in a separate box is probably not a deal maker, but a nice feature everything else being equal. Ascend is explicitly not making its own DWDM to integrate into the cabinet, but is way out in front of everyone else in delivering the capability of interfacing directly with the most popular DWDM equipment, thus eliminating the need for Sonet gear.
I thought eliminating boxes was a deal maker. Aren't these companies competing by offering total solutions? Single boxes that make efficient use of C/O real estate? And how can Ascend be way out in front "in delivering the capability of interfacing directly with the most popular DWDM equipment" if Cambrian has superior technology that is interoperable with Lucent, Bay/Nortel, IBM/ESCON, 3Com, HP, and Packet Engines/Alcatel?
From recent press releases:
* ESCON and IBM peripherals. . .
Kanata, Ontario, September 8, 1998 - Cambrian Systems Corporation, a leader in metropolitan dense wavelength division multiplexing (DWDM) solutions, will showcase its OPTera survivable ring-based system delivering the Enterprise System Connectivity (ESCON) application at the National Fiber Optic Engineering Conference (NFOEC) to be held September 13-17 in Orlando, Florida. ESCON is utilized to interface IBM peripherals and mainframe computers running mission-critical applications in the demanding healthcare and financial industries. The extensive use of ESCON makes users cautious to change existing infrastructure as the requirement for computing resources increases. Cambrian's OPTera fully supports ESCON and significantly increases the capacity of the network, while enhancing the reliability of data transmission, further demonstrating OPTera's ability to deliver survivable transport services in carrier/enterprise applications.
* Cambrian Systems and Lucent:
Kanata, Ontario, June 1 1998 - Optical networking leaders Cambrian Systems and Lucent Technologies today announced their collaboration on interoperability between their metropolitan and long-haul Dense Wavelength Division Multiplexing (DWDM) systems. Interoperable DWDM products enable service providers to build flexible, scalable, cost-effective networks.
Both companies successfully completed initial testing between Lucent's WaveStar™ OLS 40G and Cambrian's OPTeraTM Metro DWDM systems, as well as Lucent's WaveStar OLS 400G long-haul DWDM system. Cambrian and Lucent will demonstrate interoperability of their systems at SUPERCOMM'98 in both the Cambrian Booth #(8051) and Lucent Booth (#6639).,/i>
*Cambrian Systems and 3Com, Bay [now part of Nortel] and Packet Engines [now owned by Alcatel]:
Kanata, Ontario, April 27, 1998 - Networking industry leaders 3Com Corporation, Bay Networks, Cambrian Systems, and Packet Engines are joining forces to demonstrate Dense Wavelength Division Multiplexing (DWDM) and Gigabit Ethernet interoperability at NetWorld+Interop in Las Vegas, May 5-7th 1998.
This industry-first demonstration brings together the unique capabilities of Cambrian's DWDM OPTera TM (Optical Protocol Independent Transport era) and the high performance Gigabit Ethernet solutions from 3Com's SuperStackÒ II Switch 9300, Bay Networks' Accelar 1200 Routing Switch, and Packet Engines' PowerRail TM 5200 Enterprise Routing Switch.
* Cambrian Systems and HP:
SAN JOSE, Calif., November 3, 1997 -- Hewlett-Packard Company today announced a letter of intent with Cambrian Systems Corporation to develop Dense Wavelength Division Multiplexing (DWDM) for metropolitan access and inter-office networks. HP and Cambrian will combine R&D efforts to optimize and integrate their respective DWDM components and networking systems. This strategic relationship is expected to lead to network solutions that provide increased value to the metropolitan area network marketplace and service providers worldwide.
DWDM enables the transmission of many independent optical channels on a single fiber, providing a new level of affordable public network bandwidth just as Internet traffic is soaring, corporate intranets are proliferating and bandwidth-consuming multimedia applications are reaching the market. The economies of DWDM can be powerful, saving potentially hundreds of millions of dollars in additional fiber-optic plants and electronic transmission and switching equipment.
* List of articles explaining the different approaches:
cambriansys.com
"Cambrian's Optera 80 dense WDM helps net managers play it safe by building redundant rings:"
This means the mux can simultaneously send OC3 (155-Mbit/s), OC12 (622-Mbit/s) OC48 (2.4-Gbit/s) ATM (asynchronous transfer mode); FDDI; Fibre Channel; gigabit Ethernet; IBM Escon (Enterprise Systems Connection architecture); HPPI (high-performance parallel interface); SDH (synchronous digital hierarchy); and Sonet (Synchronous Optical Network) over the same pair.
A Ring of Reliability for WDM
But what really sets Optera 80 apart from the competition is its ability to be co nnected in a ring. Other WDMs can only be used to link two sites point to point. Redundancy means buying extra muxes that can be swapped in if a box fails or building a completely parallel connection over a second fiber pair. Both options are obviously expensive, particularly in networks containing multiple sites.
In contrast, Cambrian (Kanata, Ontario) built the Optera 80 so that it can be linked point to point or in a ring. Up to nine WDMs can be connected in a ring not exceeding 40 kilometers in circumference (see the figure). If a fiber segment fails or an Optera 80 goes down, data is rerouted back around the ring. What's more, this ring protection is available on a per-channel basis, so net managers can decide which traffic to protect and which to dump in case of failure. Optera also works as a passive add/drop mux, allowing data to join and leave the ring on a channel-by-channel basis.
Ascend's white paper:
ascend.com
Ascend and Pirelli:
ascend.com
I challenge you to study the two companies' capabilities and still claim Ascend's DWDM is leading the competition. In fact I challenge you to find anyone in the industry to agree with you.
Given that most fiber systems will be asked to carry more than just ATM data - (Pure IP, traditional circuit switched) - for years, particularly in the rboc space, many customers may prefer that this capability not be fully integrated into the ATM box. I believe the approach Ascend and Cisco are taking, of building in direct interfaces to a wide variety of DWDM gear is the preferable approach.
I believe NN/Cambrian will not force anyone to buy the DWDM interface if they don't want it.
Let me close by saying again that I see NN as having alot of upside potential. You have evidently spent alot of time with company management and done alot of homework. I caution you that NN has a terrible reputation on Wall Street for promising alot and delivering very little. Mr. Lutz is a step in the right direction, but he has a big job. I think you are a quarter or two away from seeing a big move in this stock, but I could be wrong.
I appreciate your concern. As for when the market will give NN its true value, I've placed my bets on sooner rather than later. If this were a chess game, I would suggest there are some moves being considered that will catch many by surprise.
I enjoy the point-counter-point and hope it continues.
Cheers!
Pat |
