Ethernet networks operating at (100+100) Gb/s per link are showing enough technical and market viability that intense development on technical specifications and component technologies is under way. However, the route to 100 Gb/s Ethernet products is still not completely clear because the market and applications for 100 Gb/s links are quite different than for 1 Gb/s and even 10 Gb/s links.
This article attempts to provide a roadmap for adoption of 100 Gb/s Ethernet in enterprise data centers, outlining the features that will affect the schedule, as well as the capabilities of Ethernet gear as it is developed.

The major opportunities for 100 Gb/s Ethernet appear to be primarily oriented toward server, rather than desktop/client, applications, and toward interconnecting various types of high-performance computing gear for technical and business analytic computing, as well as for media-oriented and web applications in content development and delivery. These applications will place requirements on the timeline and technical definition of networking gear operating at 40 Gb/s and 100 Gb/s.

One of the safest predictions about technology evolution is that higher-speed networks will be needed in the future. While the fortunes of individual networking companies have varied dramatically, the requirements for networking equipment with higher bandwidth, lower latency and more robust service has risen steadily at well over a 50 percent per year growth rate [1]

To say "we will need 100 Gb/s Ethernet" is, therefore, an easy and safe prediction. Harder, though, is to predict which of the various possible flavors of gear will be needed, in what volumes, for which markets and applications and (perhaps most importantly) when.

This article attempts to describe the potential early adopters for 100 Gb/s that will demand higher network performance and fuel the initial ramp-up of 100 Gb/s Ethernet within data centers. We do this partly by examining the types of applications that are driving higher speed links. We will also consider the related and peripheral technologies that will affect 100 Gb/s Ethernet adoption, including technologies like memory and I/O buses that impact the ability to exploit 100 Gb/s speeds, and technologies like Fiber Channel and InfiniBand, which may compete with Ethernet at the 100 Gb/s data rate.

Download a PDF of this article

ALAN BENNER is a senior technical staff member in the IBM Systems and Technology Group, in Poughkeepsie, NY, working on system architecture, design, and development of optical and electronic networks for high-performance servers and parallel systems. He received a B.S. degree in physics in 1986 from Harvey Mudd College and joined AT&T Bell Laboratories, doing photonic networks and components research until 1988. He received M.S. and Ph.D. degrees from the University of Colorado at Boulder in 1990 and 1992, researching nonlinear interactions between wavelength-multiplexed optical fiber solitons in the Optoelectronic Computing Systems Center. He joined IBM in 1992, working on RS/6000 scalable parallel systems, and has been doing R&D since then on various aspects of high-performance server networks and systems design. He has over 15 technical publications, including books on Fiber Channel and specifications for the InfiniBand architecture, and over a dozen patents issued in the United States and other countries.

PETAR PEPELJUGOSKI joined IBM T.J. Watson Research Center in Yorktown Heights, N.Y., as a research staff member in 1994 after receiving his Ph.D. from U.C. Berkeley. His research work included design, modeling, prototyping, and characterization of MMF LAN links and parallel optical interconnects. He was involved in the development of several Ethernet (IEEE 802.3z, IEEE 802.3ae, IEEE 802.3aq) standards, as well as the development of the specification for the OM3 fiber, and was subsequently recognized for his contributions. A senior member of the IEEE, he is the author or coauthor of more than 50 journal and conference articles.

RENATO RECIO is chief engineer of IBM's eSystem Networks, in Austin, TX. He has been a founding engineer and author of system I/O and network industry standards, including: Future I/O, InfiniBand, iWARP (RDMA/DDP/MPA over TCP/IP), iSER, PCI I/O Virtualization and, more recently, Convergence Enhanced Ethernet (CEE) and Fiber Channel over CEE. For the past 10 years, he has been responsible for the I/O network architecture and design of future IBM systems. He also created and chairs IBM's I/O Technical Community (IOTC), which serves the networking, education and support needs of over 750 IBM I/O folks. He has several I/O related patents and has made several external presentations. He has received three external professional awards for his contributions to the industry.