INTERESTING ROBI STEPHENS REPORT: Sorry about the format, but I don't have time to clean it...
The information contained herein is not a complete analysis of every material fact respecting any company, industry or security. Although opinions and estimates
expressed herein reflect the current judgment of the Firm, the information upon which such opinions and estimates are based is not necessarily updated on a
regular basis; when they are, the date of the change in estimate will be noted. In addition, opinions and estimates are subject to change without notice. This
Report contains forward-looking statements, which involve risks and uncertainties. The Company's actual results may differ significantly from the results described
in the forward-looking statements. Factors that might cause such a difference include, but are not limited to, those discussed in "Investment Risks." BancAmerica
Robertson Stephens from time to time performs corporate finance services for some companies described herein and may occasionally possess material,
nonpublic information regarding such companies. This information is not used in the preparation of the opinions and estimates herein. Facts and other information
discussed have been obtained from sources considered reliable but are not guaranteed. BancAmerica Robertson Stephens, its managing directors, its affiliates,
and/or its employees may have an interest in the securities of the issue(s) described and may make purchases or sales while this report is in circulation. BA
Robertson Stephens International Limited is regulated by the Securities and Futures Authority in the United Kingdom. This publication is not meant for private
customers.
The securities discussed herein are not FDIC insured, are not deposits or other obligations or guarantees of Bank of America NTSA, and are subject to
investment risk, including possible loss of any principal amount invested.
Copyright O 1998 BancAmerica Robertson Stephens
THE FIRST MILE - RELEASE 1.1
"Every journey starts with the first mile."
- Anonymous
Welcome to the premiere issue of The First Mile - an e-mail periodical for the distribution of
new thoughts, information and trends relevant to the data communications industry. This
newsletter will serve to keep subscribers abreast of the latest happenings in our coverage universe
with a focus on technology trends and development.
Is this the Year of ATM?
We have believed for a long time that ATM would show itself useful to carriers and
enterprises as a backbone transport. However, ATM has floundered for many years, even when
each year was touted as being, finally, "The Year of ATM." New initiatives such as packet-over-SONET
(PoS) have taken interest away from ATM, as they allow TCP/IP to be carried over
SONET without the need for ATM as an intermediate switching fabric. Additionally, it has been
demonstrated that using an ATM switching fabric in the core of the Internet would not work since
the fabric would not be able to keep up with the demand of setting up and tearing down many
short-lived streams.
That being said, it is becoming clear to us that ATM will be used in the core of these large
networks at the very least as a transport. While some complain of the ATM "cell tax" (nearly
10% of an ATM cell is header information), it has become clear that the "router tax" (every hop
where traffic is routed introduces additional delay called latency) is a heftier sum. Carriers are
willing to pay (in dollars and bandwidth) for deterministic networks, and the ATM cell tax gets
cheaper as ATM speeds increase.
One of the more striking reasons we have heard supporting ATM versus PoS in the core is the
fact that the majority of data traffic is still carried on Frame Relay. While ATM effortlessly slices
Frame Relay into ATM cells and back again, PoS has no facility to carry Frame Relay traffic and
then put out native Frame Relay at the other end. Finally, ATM was designed from the ground up
to offer QoS, and TCP/IP may never evolve a scheme that will work across disparate networks.
While we believe that ISPs will continue to deploy IP-centric networks where QoS plays a
smaller role and we may see an extended period of time where two distinct networks exist
separately, we believe that over time these networks will converge as large carriers deploy
2
scalable networks based on ATM with the capacity and Qos to satisfy the requirements of all
forms of data, voice, and video.
Recent evidence of the improvement in the ATM market outlook includes:
ú Williams currently has an 11,000 mile network and recently announced plans to purchase
$150 million of Ascend (ASND $28, BUY) ATM equipment (~100 chassis). The company
plans to expand its network to 18,000 miles by the end of 1998 and 25,000 miles by the end
of 1999, and it expects to spend $1 billion in the process. While most of the $1 billion will
not go toward core switching products, it still offers an impressive potential.
ú Bell Atlantic recently announced that it would be offering native ATM service in the heritage
NYNEX territory with a rollout of 36 Ascend ATM switches. The company will also expand
the existing ATM service in its heritage territory.
ú AT&T just announced its plan to deploy more than $8 billion in 1998 to "future-proof" its
network. In addition, the company is planning to add approximately 200 edge switches to
meet the growth in demand of Frame Relay, ATM, and IP services.
ú The founders of MFS have begun a new network called Level 3 Communications and plan to
take its 6-city presence to 60 by 2001, ramping up to 20,000 route miles in the process.
ú Additionally the CTR group has a $14 billion plan called Project Oxygen that will slather the
globe with 198,400 miles of fiber using ATM switching at the core. (Wow!) While these
projects may be a bit pie-in-the-sky, we believe they underscore carriers' intent to deploy
high-capacity networks based on ATM.
ú Carriers are becoming more interested in selling bandwidth on a wholesale basis, making the
ability to partition and guarantee bandwidth with ATM essential selling points.
Based on this reasoning, we upgraded Ascend to Buy from Long-Term Attractive and raised
our estimates based on the fact that we see ATM network deployment accelerating and an
acceleration in demand for the company's core switching products_primarily the CBX500 and
recently introduced GX550 ATM products from Cascade.
Recent discussions with carrier customers have indicated that demand for core ATM products
is beginning to accelerate and that the GX550 is potentially a "killer product." We believe that
Ascend's CBX500/GX550 could emerge as one of the few critical components within next-generation
carrier networks.
In the same vein, we recently upgraded TranSwitch (TXCC $9-7/8, BUY) to Buy from Long-Term
Attractive. TranSwitch is a fabless semiconductor company that supplies advanced
integrated circuits (ICs) used in broadband telecommunications applications.
The company is a provider of mixed signal, integrated circuits used in systems based on the
SONET/SDH broadband standard and is emerging as one of the few vendors in the market for
ATM ICs. During the past 12 months the company has announced key system design wins with
leading telecommunications and networking vendors such as Cisco Systems, Newbridge
Networks, Ascend, Ericsson, Bay Networks, and Tellabs. As such, we believe the company is
well positioned to benefit from the accelerating deployment of large-scale SONET and ATM
networks.
In our next issue we plan to discuss high speed access technology, including traditional analog
modems, xDSL (Digital subscriber Line), and cable modems.
BancAmerica Robertson Stephens maintains a market in the shares of Ascend Communications, Cisco
Systems, L.M. Ericsson, Transwitch Corporation and has been a managing or comanaging underwriter
for Ascend Communications and Transwitch Corporation within the past three years.
Related Web Links:
Ascend Communications - ascend.com
TranSwitch Corporation - transwitch.com
CTR Group - oxygen.org |
