Fast Ethernet

Ka Kavitha V Updated 08 Oct 2026
7 min read ·Lesson 9 of 26

Fast Ethernet

As computer networks evolved, the need for faster communication between devices became increasingly important. Early Ethernet networks operated at 10 Mbps (megabits per second), which was sufficient for basic tasks such as file sharing and email. As organizations adopted multimedia applications, larger databases, and client-server architectures, network traffic grew well beyond what 10 Mbps Ethernet could comfortably handle.

To address this, Fast Ethernet was introduced as a major upgrade to traditional Ethernet, increasing network speed from 10 Mbps to 100 Mbps — ten times the bandwidth — while remaining compatible with existing Ethernet standards. It became one of the most widely adopted LAN technologies of the 1990s and early 2000s.

What Is Fast Ethernet?

Fast Ethernet is an enhanced version of Ethernet that supports data transmission speeds of 100 Mbps. It was standardized by the IEEE (Institute of Electrical and Electronics Engineers) as IEEE 802.3u in 1995.

Traditional EthernetFast Ethernet
Speed10 Mbps100 Mbps
Improvement—10× faster

Fast Ethernet uses the same Ethernet frame format and core networking principles as traditional Ethernet. That compatibility was the whole point: organizations could get a tenfold speed increase without redesigning their networks from scratch.

Key Terms to Know First

A few terms are worth pinning down before going further.

  • Ethernet — the most widely used technology for wired LANs. It defines how devices communicate, how data is transmitted, how access to the shared medium is controlled, and how collisions are handled. Ethernet organizes data into units called frames.
  • Data Transfer Rate (DTR) — the speed at which data moves between devices, measured in bits per second (bps), kilobits per second (Kbps), megabits per second (Mbps), or gigabits per second (Gbps). A Fast Ethernet link can theoretically move 100 million bits every second (100 Mbps).
  • Bandwidth — the maximum amount of data a connection can carry in a given time, often compared to a road: a wider road (higher bandwidth) carries more traffic before congestion sets in. Fast Ethernet provides 100 Mbps of bandwidth, compared to 10 Mbps for traditional Ethernet.

Why Fast Ethernet Was Introduced

Ethernet was originally developed in the 1970s at Xerox PARC, running at 10 Mbps over shared coaxial cable. By the early 1990s, organizations were running shared databases, multimedia applications, email systems, network printing, and client-server applications — and 10 Mbps Ethernet had become a bottleneck, producing slow file transfers, network congestion, and reduced productivity.

In 1995, the IEEE released IEEE 802.3u, defining Fast Ethernet at 100 Mbps. Its major improvements were:

  • A tenfold increase in speed (10 Mbps → 100 Mbps)
  • Support for better cabling, both twisted-pair and fiber (covered below)
  • Backward compatibility with existing Ethernet equipment and cabling practices

Because Fast Ethernet preserved Ethernet's frame format and most of its MAC-layer behavior, organizations could upgrade their networks incrementally rather than replacing everything at once — a major reason for its rapid, widespread adoption.

How Fast Ethernet Works

Fast Ethernet keeps Ethernet's basic structure while improving MAC-layer operation and introducing new options at the Physical layer.

Media Access and CSMA/CD

The MAC (Media Access Control) layer controls how a device gains access to the shared network medium. Fast Ethernet preserves Ethernet's MAC-layer design, which is why older and newer Ethernet devices can coexist on the same network.

When operating in half-duplex mode — sharing a hub or a single cable with other devices — Fast Ethernet uses the same channel-access method as traditional Ethernet: Carrier Sense Multiple Access with Collision Detection (CSMA/CD). A device using CSMA/CD:

  1. Listens ("carrier sense") to check whether the medium is currently idle.
  2. Transmits if the medium is free.
  3. Detects a collision if another device transmitted at the same time.
  4. Waits a random backoff interval and retransmits if a collision occurred.

Full-Duplex Operation

Fast Ethernet's most important practical improvement is support for full-duplex operation over switched connections. When each device has a dedicated link to a switch port, rather than sharing a cable with other devices, it can send and receive simultaneously, with no possibility of a collision. Because there's no shared medium to contend for, full-duplex Fast Ethernet doesn't use CSMA/CD at all — CSMA/CD is only needed on genuinely shared, half-duplex connections. (See the lesson on transmission modes for more on how simplex, half-duplex, and full-duplex communication work.)

A device connected in full duplex can, for example, upload and download data at the same time without waiting for the line to clear.

Physical Layer Standards

Fast Ethernet defines several physical-media standards, collectively known as 100BASE-T:

StandardMediumMax segment lengthNotes
100BASE-TXCat5 (or higher) twisted-pair copper, 2 pairs100 metersBy far the most common Fast Ethernet standard; used in office, school, and home LANs
100BASE-FXFiber optic cableSeveral hundred meters to a few kilometers, depending on fiber typeImmune to electromagnetic interference (EMI); used for longer runs and building-to-building links
100BASE-T4Cat3 (older) twisted-pair copper, 4 pairs100 metersDesigned to reuse older telephone-grade cabling; largely obsolete today

To reach 100 Mbps over these media, Fast Ethernet also introduced new line encoding:

  • 4B/5B encoding — every 4 bits of data are mapped to a 5-bit symbol before transmission. The extra bit per symbol improves clock synchronization between sender and receiver and reduces transmission errors.
  • MLT-3 signaling — used specifically by 100BASE-TX, this three-level signaling scheme reduces the signal's electromagnetic emissions, allowing reliable 100 Mbps transmission over ordinary twisted-pair copper cable.

Common applications:

  • 100BASE-TX — office LANs, schools and colleges, small businesses, and home networks. Computers connected to an office switch with Cat5e cabling typically run 100BASE-TX.
  • 100BASE-FX — campus networks, industrial environments, and building-to-building connectivity, such as a university linking two buildings across campus with a fiber run.

Fast Ethernet in Network Topologies

Fast Ethernet can run over different network topologies, though not all are equally practical:

  • Star topology — by far the most common Fast Ethernet deployment. All devices connect to a central switch, which makes troubleshooting straightforward, improves reliability, and simplifies expansion. Most office networks use this arrangement.
  • Bus topology — technically possible on shared coaxial segments, but rarely used today due to frequent collisions, poor scalability, and difficult troubleshooting.
  • Hybrid topology — combines multiple topology types for flexibility, often used by larger organizations connecting several star-wired segments together.

(See the lesson on network topologies for a full discussion of these topology types.)

Advantages of Fast Ethernet

  • Backward compatibility — can operate alongside traditional 10 Mbps Ethernet devices, simplifying gradual upgrades.
  • Cost-effective — delivered strong performance at a lower cost than contemporaries like ATM or FDDI.
  • Easy to deploy — reused existing Ethernet knowledge, cabling practices, and network management tools.
  • Improved productivity — faster file transfers and application access for end users.
  • Wide industry support — adopted broadly across networking vendors, making it a de facto standard for over a decade.

Limitations of Fast Ethernet

  • Limited speed — 100 Mbps is far below what Gigabit Ethernet (1 Gbps) or 10 Gigabit Ethernet (10 Gbps) can provide, and is no longer adequate for most modern workloads.
  • Scalability challenges — large, data-intensive networks quickly exhaust 100 Mbps of shared or per-port bandwidth.
  • Distance restrictions — 100BASE-TX is limited to 100 meters per cable segment; longer distances require additional switches, repeaters, or a move to fiber (100BASE-FX).
  • Legacy hardware costs — maintaining old Fast Ethernet switches and network cards becomes increasingly inefficient and expensive as newer standards take over.

Real-World Applications

Fast Ethernet was widely deployed across many kinds of networks:

  • Data centers — connecting servers, storage systems, and switches, before the move to Gigabit and faster standards.
  • Education — providing internet access, online learning platforms, and shared resources in school and university computer labs.
  • Healthcare — transferring X-rays, CT scans, MRI images, and electronic health records between systems, improving collaboration among medical staff.
  • Telecommunications — carrying backhaul traffic and data between network nodes.
  • Industrial automation — connecting PLCs (Programmable Logic Controllers), sensors, and monitoring systems for real-time production monitoring.
  • Surveillance — transmitting video feeds from IP cameras to monitoring and recording systems.

Today, Fast Ethernet has largely been superseded by Gigabit Ethernet and faster standards, but it remains a foundational technology for understanding how modern wired Ethernet networks evolved.

0 Comments

Reviewed before they appear

No comments yet.

Computer-Network
Ask about this post
AI Ask about this post

Ask questions about Fast Ethernet and get answers drawn from it.

Signed-in readers only.