Good Article on Avici (Nortel owns 15-20%) and terabit routers.
Are You Aware of Avici?
Cracking Open the Market for High-End Routers
Jul 24, 2000, 10:41 am EST
By Yusuf Haque
NEW YORK (123Jump) On January 06, Enron Communications (ENE) startled
the telecommunications world by announcing the company’s intention to test
Avici routers for deployment in its broadband network - heralding a significant
change in the market climate for high-performance routers. The announcement
signified two things. First, it affirmed that the new requirements of systems
providers across-the-world would necessitate an increase to terabit capability.
Second, it confirmed that the ocean of photonic pulses, in the deep core of the
network, would now be ushered in and out by higher capacity routers - the likes
of which Juniper Networks (JNPR) and Cisco Systems (CSCO) do not currently
have to offer.
July 27, Morgan Stanley Dean Witter will close the allocation books for the initial
public offering of Avici Systems. The sale of 6 million shares to the general public
is expected to price within a range of $18.00 to $20.00, raising $114,000,000 for
the young company already 20% owned and backed by Nortel Networks (NT).
The company filed May 18, but has been cautious about bringing its shares to
the market. The reason being the current market environment - although
by-and-large friendly to networking companies - no longer disposed to look kindly
upon start-ups with meager revenues and small customer bases. Nonetheless,
the story of the “terabit trendsetter” is likely to generate one of the largest
amounts of block "buy" orders this IPO market has seen for some time.
A Sought After Solution
Today's communications networks apply optical networking technology to
alleviate and replace many of the hindrances that traditional copper-based
networks pose for the future of communications technology. Not only have optical
networks been introduced in a cost effective manner, they have also opened the
door to wide deployment of broadband Internet access - the demand for which is
forecast to grow at a compound annual growth rate of 77% for the next three
years.
Significant advances in optical technology - such as soon to be introduced
Lucent (LU) and Nortel DWDM transmission systems which belt data at speeds
of OC-768 or 40 Gbps through fiber strands - have outpaced advances in packet
switching and routing technology; a market that Cisco was once largely
responsible for pioneering. Large bottlenecks have arisen in communications
networks at the points where optical core meets packet cell switches - intended
to route data more efficiently.
The solution to this problem has been, for systems providers like Williams
Communications (WCG) and AT&T (T), to deploy super-fast routers which can
forward data at optical speeds. The resultant fact that optical networks will
become inevitably more prevalent in the telecommunications backbone, will
cause the market for IP backbone routers to flourish.
A Challenge to Dominance
The market for high-end routers is currently divided in an 85 to 15 percent
distribution, between Cisco and Juniper - an addressable market-size that
currently stands at $1.6 billion, headed for $5.5 billion by 2003. The two
companies' current product lines, however, are no longer the most powerful
solution on the block. Additionally, as total network traffic zooms towards a
bewildering 16 million terabits of data per month in 2003 (RHK), CLECs and ISPs
are rattling their knives and forks in demand for more bandwidth.
In March of this year Juniper introduced its largest and most powerful router, the
M160 next generation solution with OC-192 interface. However, systems
providers' experience with the M40 router has been that a 12-14 month life cycle
is too short - the product seemingly comes a little too late to satiate providers’
demand for increased processing capability in the optical network core. The new
generation of Juniper products, therefore, will most likely be used toward the edge
of the IP core.
Cisco - employing marketing tactics that only a multi-billion dollar company can
afford - boldly announced earlier this year its 12000 terabit system. Essentially
an array of its 12016 chassis connected by a terabit switch fabric - it is scorned
by existing customers as being wasteful of precious port interfaces and
expensive.In addition to this, no one is sure whether Cisco can actually deliver
the solution - due to the fact that the aforementioned switch fabric hasn't actually
been developed yet. The company acquired Growth Networks recently, (I might
add, subsequent to declaring the company’s new capability), which may bear the
fruits of the missing link - however until then, customers may well have to see the
terabits pulsing before they believe it.
Ahead of the Hordes
Avici was the first company to introduce the concept of the terabit router back in
1997.A product of months and months of development, its TSR terabit router, was
unveiled at Networld+Interop on May 11, 1999. Perceived as an industry
milestone, the switch was demonstrated, and will be sold, in conjunction with
Nortel's OC-192 transport node as the Versalar TSR 45000. The TSR is capable
of supporting over 8,000 OC-3c or 560 OC-192c connections, but can mix and
match within various speeds. The high-resiliency control software used in the
system incorporates thin-layer 1 and 2 in order to easily integrate and support
frame relay, ATM (Asynchronous Transfer Mode) and SONET (Synchronised
Optical Network).
IronBridge Networks, Charlotte's Web Networks (Owned by MRV (MRVC)),
Nexabit (Owned by Lucent) and Tellabs (TLAB), each claim to have a terabit
router. The potential market will evolve accordingly, over time, as each player
finds its alcove of customers. Product differentiation, in the case of routers, will
include support for various interfaces, or as Avici has introduced, a
self-perpetuating switch fabric. The Charlotte's Web router, close on the heels of
Avici, can gurantee QoS (quality of service) by supporting TDM connections,
which prioritize signals with respect to their medium. For example, voice signals
are usually given priority because of the timeliness and sensitivity of the signal.
IronBridge has noted its intention to differentiate itself by using an all-optical
switch fabric - as opposed to common copper-based methods.
Avici's TSR also stands out among its peers due to its hardware based
Application Specific Integrated Circuits (ASICs) that dramatically continue where
software based solutions, unable to keep-up with internet traffic loads, cease.
ASICs enable prioritization, based on requested QoS, without compromising
speed and efficiency - promising the possibility of building one common optical IP
infrastructure, as opposed to the various layers that are prevalent now.
The market for routers is not unique in networking equipment, in that the success
of any one company will depend on who is working with you and who is giving
you support. In telecommunications, even the largest players like Siemens can
wait around for contracts for months. Avici seems to have addressed this issue
adequately. Working with Nortel, Corvis, Williams, Deutsche Telecom (DT),
Enron and Amber Networks, has assured the communications world of its
intention to address the needs of its networking contemporaries. In stark contrast
to Cisco, whose terabit solution implies dramatic resource waste for its
customers, Avici's is scalable, effective and provides a cost-savings of
approximately 50% per port.
The young company has made significant headway into the market, as evidenced
by systems provider readiness to test and buy their router. However, to be able to
penetrate the space to the same extent, Avici should take lessons from Juniper's
example of ensuring interoperability with “gorillas” such as Cisco. In fact, one of
the prime reasons behind the success of the Juniper routers has been the
JUNOS internet software which is fully interoperable with Cisco's IOS - the most
widely deployed operating system today. If Avici can surmount this significant
barrier to entry, its routers could be installed at critical points within any network
core.
Another key issue will be manufacturing capacity and its management. It is vitally
important in today's networking world to satisfy demand, and not keep providers
waiting. Demand emanates from the consumer, is funneled through to the ISP,
like Concentric (CNCX), is then aggregated at the systems provider, like
Williams Communications. If the manufacturers of the systems contained within
a network are not timely, this delay will likely result in a loss of customers.
Marconi is an example of a company whose books are thoroughly clogged - and
are sacrificing company reputation as a result.
The head-on competition in the space will, in the short-term, come from Lucent's
Nexabit. However, Avici has several convincing advantages.First, Avici was able
to bring its TSR to the market well in advance of Nexabit, who started shipping
only within the last two months. Second, Nexabit has already witnessed the
emergence of software glitches in its product. Third, Lucent has had considerable
trouble integrating Nexabit, having also suffered the loss of the start-up's CEO.
What many investors will be considering is Avici's financial performance to date -
one reason why it has delayed the first trading day until now. Revenues for the
first quarter of this year were $504,000 with a gross margin of 14.8%. Net losses
totaled $16.6 million for the same period. Crucial to understanding Avici, is that
its product was only ready for testing at the end of 1999, and conclusive evidence
indicates that Avici's TSR is the terabit router to beat. Enron Broadband Services
and Williams communications have agreed to future minimum purchases totaling
$45 million at least, through 2001.
With approximately 45 million shares outstanding after the offering, the
company's market capitalization could easily reach $1 billion on the first day of
trading. The crucial variable to watch on first-day trading will be volumes. The
allocation books have been extremely tight, as indicated by MSDW, and a
number under 10,000,000 will be highly positive if purchasing shares in the
aftermarket. Consider also that optical networking IPOs are often highly valued,
and the investor temptation to flip on the first day may be too much to resist if the
price begins to skyrocket. This, therefore, may be a highly volatile stock for the
first few days subsequent to the pricing.
In summary of the points presented above, Avici Systems may be the next
benchmark product in routing/switching technology. The company was clearly
able to predict the necessity for such a product as long as three years ago, when
network traffic was a fraction of that in existence today. Chances are, Avici is
already thinking about the future needs of carriers, and this offering will aid them
in cultivating the company’s answer to the next question - before it has even been
asked. The next two years will, at the very least, be intriguing - watching how
Juniper and Cisco hold their ground against the determined entrants, who
promise to blow apart this tightly-clinched market. (Previous Stories)
(c) Copyright: 2000 123jump.com, Inc.
123jump.com |
