iceberg you wrote:
<<This is the best news for Ancor that I HAVE SEEN IN AGES!!!
With Brocade's SilkWorm switch currently in the top 15-20 OEM's
evaluation or qualification cycles
Guess what other switch is in the exact same boat and is better in
price, performance, and design. Ancor will win their share and this
stock will quadruple>>
I find it interesting that six months ago Vixel signed a $50 million deal to supply FC hubs to Compaq over a 2-3 year period, yet things still look very rosy for Gadzoox, a direct competitor to Vixel. Could it be that a rapidly expanding FC market allows more than one FC hub vendor to survive and thrive? If Gadzoox had been a public company 6 months ago, would their shareholders be screaming bloody murder for allowing Vixel to "steal" a $50 million contract?
On a separate note, the past several days "revelations" about Brocade's ability to obtain financing and their most recent $6.5 million quarter should not be a big surprise to any long time followers on this thread. In fact, I would be surprised if a big chunk of it didn't come from Sequent, considering that when Ancor still had the deal, they were expecting at least $6 million from Sequent in 1997. What is unclear is how much of these revenues were channel sales ( Prisa, McData, OEM evaluation units ) as opposed to end users? Assuming that Brocade's numbers translate to actual end users, this bodes well for the both the market demand for Gigabit solutions as well as FC switches. The biggest questions facing any emerging data communications technology is:
1) the acceptance of the technology ( FC ) over other technologies ( ATM, GE, FDDI, SCSI, SSA etc..),
2) whether real world applications require Gigabit speeds ( now we have proof ) and
3) the demand for switches in the FC architecture ( as opposed to just adapters and hubs ).
Need I remind people that skeptics have repeatedly posted their concerns about the future potential of the FC market as well as the need for FC switches? The biggest hurdle for development stage companies early in the product life cycle is market adoption, not who wins which account. Amati's biggest challenge has always been IF and WHEN ADSL will be deployed on a large commercial scale, not who wins NYNEX or GTE on an initial basis. This is one of the biggest reasons why I have always been so skeptical of NUKO. Where and how big is the MARKET for NUKO's products? Where did NUKO's partnerships with Cisco, ADCT and ORCL get them? The OTC Bulletin Board? ..."It's the Market stupid."
The Reality is that if any OEM required a Class 2/3 switch running at full gigabit speed to ship/interoperate with other FC components in 1997, then they had no choice but to buy from Brocade. The Reality is that Ancor's new MKII was simply unavailable for testing until the second half if 1997. Those who could not wait for Ancor, ie SQNT/DEC moved ahead with productization. WHY didn't Ancor have MKII ( Class 2/3 at full gig speed at a competitive price ) ready for prime time when Sequent and DEC needed it? Four reasons: 1) Terry Anderson ( Class 1 zealot ) 2) Dale Showers ( asleep at the switch and golfing in Florida all winter long ) 3) Steve O Hara ( self-explanatory ) and 4) Sequent ( the systems integration problem caused Ancor to redeploy engineers from working on MKII to working on Sequent which probably delayed MKII by at least 2-3 months ). That's history. Now it's a horserace between a few competitors.
I would encourage anyone out there to state why they think Brocade will "lock up" all the FC switch OEM business and "dominate" the category. Ed and Rob, take a crack at it. For the investor/trader with at least a 6 month timeframe, why does ANCR look bad from a risk/reward perspective going forward? What is the basis for any alleged "technological superiority" over Ancor that will entrench Brocade as the "Intel/Microsoft/Iomega of Fibre Channel switching? Does anyone have any data/evidence of any Brocade cost advantage over Ancor? What sustainable/long term competitive advantage does Brocade have over Ancor, Arxcel and McData??? The last time I checked, although Cisco is the clear frontrunner in networking, 3 Com, Bay and Cabletron are still in business as are many others <g>.
Do people dispute the cost/margin benefit of Ancor having 4 ports per ASIC? Obviously, many on this thread are confusing end user list price with the real contract prices quoted to OEM's.Suggest they look at a recent MRVC press release on one of their new fast ethernet switches that touts the cost/performance benefits of integration on the ASICS:
CHATSWORTH, Calif.--(BUSINESS WIRE)--Nov. 19, 1997--MRV Communications (NASDAQ:MRVC) subsidiary NBase Communications announced today a new member of its Fast Ethernet switch family, representing a breakthrough in price per port and functionality.
The new NH 2024T is a highly integrated, ASIC based auto- negotiating switch supporting both 10 Mbps (megabit per second) and 100 Mbps on each of its 24 ports. The base unit which supports 24 10/100 ports can accommodate one uplink module, with eight (8) 10/100 auto-negotiating copper ports, or four (4) 100 Base-Fx ports, providing fiber optic transmission capability up to a distance of 100 km. Additional modules providing either Gigabit Ethernet, ATM or FDDI interfaces will be available during the first calendar quarter of 1998.
The 24 port base unit is list priced at $4,750, providing a price per port of $198. An uplink module with eight (8) additional 10/100 Mbps ports costs $1,995. A fiber optic four (4) port module is priced at $2,495.
Due to the highly integrated ASIC technology and four gigabit multibus architecture, the NH 2024T achieves a maximum throughput of 3,600,000 pps (packets per second).
Additional features include:
-- IP Switching Capability (each device is DirectIP enabled)
-- IP Based VLAN (First quarter 1998)
-- Web Based Management (First quarter 1998)
-- Port trunking provides a 1.6 Gpbs fast pipe between switches
-- Full support for SNMP and RMON, on a per port basis
-- Comprehensive Console Support via RS 232 or Telnet Connection
-- Multi-layer, policy based, support for VLAN's
-- 12 MB of buffer memory and support for over 8,000 MAC
addresses, per VLAN
-- Spanning Tree Algorithm
Noam Lotan, President and CEO, commented: "High Density 10/100 Ethernet Switching is essential for the seamless integration of Fast Ethernet and legacy 10 Mb Ethernet networks. We are offering this functionality at a breakthrough price. The added functionality of IP Switching, Gigabit, ATM connectivity make this product both a straight cost effective solution today, with future proof connectivity to more complex networks. This is a very compelling proposition". |
