To: craig crawford who wrote (11956 ) 12/9/1997 11:44:00 AM From: gfr fan Read Replies (4) | Respond to of 45548
Ramsey Su and Co: There was recent discussion re: if networkers were struggling with a slow down domestically and hoping that Asia would pull them out. Someone hypothesized (I think Craig) that the domestic market was in a glut stage. From my vantage point this is not true. Domestic demand for the enterprise is strong, and appears to be getting stronger, as the industry is about to enter the following upgrade cycle: Most LAN connections are still 10Mbps shared (the old shared hubs that were Ctron's and Synoptics' bread and butter). The last two years has seen a move to 10Mbps switched. Now the game is changing to 100Mbps switched - these are the desktop connections. What makes this possible is advances in silicon - ASICs. Networkers shoot for the "sweet spot" - $250 - $350 per port. This used to be the price of 10Mbps shared - it will be the price of 100Mbps switched in 1998. This represents a two orders of magnitude increase in desktop bandwidth. This flows perfectly with most NICs now being 10/100. This will allow enterprises to start offering true video and voice apps of their LAN. Who is best positioned to lead this market on both sides of the wire - COMS already has the lead at the desktop, and will have the strongest suite of products in the wiring closet in 1998. Now, once the desktop and wiring closet infrastructure is upgraded, enterprises will need to upgrade their backbones - that currently are mostly collapsed router backbones. Their is absolutely no way that any vendor's current router can handle and aggregate the types of bandwidth that will be coming from desktops and wiring closets. What does this mean? A new round of buying in the backbone of next generation routers that are ASIC/hardware based and are starting to hit the streets. Who is the leader in this field? 3Com is winning multiple industry awards for the CoreBuilder 3500/9000. Once the LAN is upgraded, enterprises will then move to the WAN. Traditional WAN links of 1.5Mbps and under will be dinosaurs and there will be another round of upgrades to ATM and SONET services at speeds of 10Mbps and greater. There will be large orders for both enterprise WAN access gear and Telco infrastructure. The leaders here? CSCO and NN on the telco side, and CSCO and COMS on the enterprise side. Finally, remote users who access enterprise networkers will need to be upgraded from 28.8/56k dial up connections. What will the technology be - cable modems? XDSL? ISDN? Still too early to tell, but the upgrade cycle is inevitable. The demand for this scenario is here now, and customers are placing orders laying the first pieces of this story. The technologies that are starting to be released as described above will accelarate demand, assuming the economy stays in check, as legacy infrastructures will be badly outdated and under equipped. I haven't even touched on the internet buildout, which will be equally huge, or on the introduction of networking products via NICs, modems, hubs, and routers into homes and small businesses. Or on the infrastructure programs that will certainly take place in India, China, Brazil, Indonesai, SE Asia, Mexico, etc. As these countries increase their PC adoption networking IS going to happen. The challenge for the networkers is to be in a position to ride these upcoming growth waves. I think there are two that have placed the right bets and are in a position to run away with the industry while the rest consolidate or die. Yes there will be bumps along the way, but networking will still be a high growth/high demand segment.