The trade press is touting Gigabit Ethernet, a third generation Ethernet LAN technology, as the solution to packet versus switched data network problems facing the IT and communications industries. Positioned as a direct competitor to switched ATM technology in the high-speed LAN market, Gigabit Ethernet-a broadcast solution-is expected to grow to almost a $1.5 billion market by 2001. A technology plan recently developed by Technology Futures, Inc. for a major university included an in-depth review of ATM and Gigabit Ethernet as part of a critical examination of the various available transport options.
Our conclusion: If you need a short distance (less than 50 feet!), high-performance network, Gigabit Ethernet may be the answer, but serious limitations to the existing technology make it unsuitable for most LAN/WAN applications. There are a number of issues concerning Gigabit Ethernet, such as standards, interoperability, and product availability, which must be addressed before it can be a market success. In the interim, there are a number of existing and emerging technologies that may well displace Ethernet technology. We'll address each of these issues briefly.
No matter which signaling scheme is selected by IEEE, the gigabit technology will be very complex, and it is important to remember that implementation complexity and network robustness do not always travel in pairs. A new and emerging Internet Engineering Task Force (IETF) standard for IP called Multi-Protocol Label Switching (MPLS) could be the key standard that merges the best of both IP and ATM. This protocol will allow IP or ATM traffic to be efficiently routed over either or both IP and ATM networks, maintaining the better routing protocols of IP and quality of service capability of ATM.
Category 5 patch cable, which is four-pair stranded PVC cable, is the industry standard for connecting networked workgroups and desktops. However, it is a two-pair specification and was never intended to supply signaling on more than two pairs. Ethernet LANs running at 10 Mb/s or 100 Mb/s (Fast Ethernet) use only two pairs-one in each direction-and are therefore subject to the crosstalk from only one pair at each end. A Gigabit Base-T device, on the other hand, transmits on all four pairs and in both directions simultaneously! As a result, it is subject to crosstalk from three pairs at either end.
In addition, the power consumption of a gigabit Base-T device is expected to be in the range of three to four watts. Put another way, it is a transmitter with the power of CB radio. Category 5 patch cable is not shielded, and there is great concern that running that much wattage over unshielded cable will create problems with other equipment. Technologies such as wireless LANs and analog cellular may not work at all around Category 5 cable running Gigabit Ethernet because of the interference.
However, to meet the growing demand for greater bandwidth, all of the new increases in usable bandwidth of unshielded twisted pair cable will need to rely on using all four pairs as signal pairs. There are a number of emerging technologies which can solve this problem including Category 5 enhanced cable, Category 6 cable which already exists in Europe, Category 7 cable, and fiber optics. This will surely change the way we design distribution networks. Gigabit speeds are three orders of magnitude faster than 10-Base-T and is part of what got the networks to where they are.
Asynchronous Transfer Mode (ATM) is a cell-switched architecture promising variable bandwidth on demand from 51 Mb/s to more than 2.4 Gb/s, quality-of-service guarantees, virtual network support, and many other features. It is a mature technology which fully supports voice and video features. Driven by collaborative computing applications such as graphics, multimedia, and video, its advantages are its ability to multiplex voice, video, and data over the same infrastructure and the fact that it facilitates a homogeneous WAN/LAN infrastructure. If standardized and implemented correctly, this standard promises to be the LAN and WAN choice for the foreseeable future.
Another way of looking at this is to examine the ATM demand at high-performance mainframe sites. As an early indicator, it is estimated that, by the end of 1998, ATM probably will most likely carry the majority of WAN traffic at leading edge, high-performance mainframe customer sites. We project that ATM will continue to dominate the telecom network backbone, despite competition from Ethernet technologies.
Although Ethernet and ATM are the dominant players in the high-speed data services field, there are some new technologies emerging that are likely to be strong contenders. These include HiPPI, Fibre Channel, and FDDI.
If you're looking to increase you network performance, there are several ways you can hedge your bet. With the exception of Fast Ethernet, ATM, and FDDI, the emerging network technologies mentioned above are still under development and may not survive. Therefore, you may want to wait until the network war settles down before making a choice.
If you believe that TFI could be of assistance to your organization in identifying and evaluating new product opportunities and in developing strategies for bringing these products to the marketplace, please contact John Vanston, Larry Vanston, David Smith, or any other member of the TFI consulting staff. (800) 835-3887
Copyright © 2000, Technology Futures, Inc.
All rights reserved. No part of this white paper may be reproduced in any form or by any means, other than short quotes to be used in reviews, without express permission from the publisher.