Decoding Ethernet Application Names: The Ultimate Technical Guide

Ethernet is the most popular communication protocol for everything from local and wide area networks (LANs and WANs) to metropolitan area networks (MANs). In today’s data centers, it also handles critical switch-to-switch and switch-to-server communications. 

To support these diverse systems, the IEEE 802.3 Ethernet Working Group develops application-specific standards for an array of media, including balanced twisted-pair copper cabling, multimode and singlemode optical fiber, twinaxial cabling, and even printed circuit board (PCB) copper traces for electrical backplanes and chip-to-chip communications. 

To keep this broad ecosystem organized, the Ethernet Working Group uses a shorthand naming system for different applications. Although these names may seem cryptic at first, they act as a quick reference. Once you understand the format and suffixes, you can tell an application’s speed, media, distance, and signaling type at a glance.

It’s All About the BASE

The IEEE relies on a standardized naming format for Ethernet applications:

[Speed] + [Baseband] + [Medium/Technical Detail]

The term BASE stands for baseband signaling, which is the foundational transmission method for Ethernet. Baseband transmits signals bidirectionally as discrete digital pulses, using the entire frequency range and bandwidth of the cable.

The prefix preceding “BASE” indicates the network speed: 

• Megabit Speeds: Represented by a standalone number. For example, 10BASE, 100BASE, and 1000BASE indicate speeds of 10 Mb/s, 100 Mb/s and 1000 Mb/s.
• Gigabit Speeds: Represented by a number followed by a “G.” For example, 10GBASE, 100GBASE, and 400GBASE indicate speeds of 10 Gb/s, 100 Gb/s, and 400 Gb/s. 
• Terabit Speeds: To support skyrocketing bandwidth demands, the IEEE introduced a “T” to denote Terabit speeds. For example, 1.6TBASE indicates a speed of 1.6 Tb/s, and 3.2TBASE indicates a speed of 3.2 Tb/s.

Deciphering the Suffix 

Following the speed and the “BASE” term is a hyphen and one or more letters indicating the Ethernet application type. These suffixes act as a technical spec sheet, revealing information such as the type of cable media, wavelength, and physical reach. The following are the most common suffixes for Ethernet applications:

• T (Twisted Pair): Runs on twisted-pair copper cabling, such as 4-pair Category 5e, Category 6, and Category 6A. Examples include 1000BASE-T for 1000 Mb/s operation or 10GBASE-T for 10 Gb/s operation. Note that if the application runs over a single twisted pair, it is denoted by a “1,” as in 1000BASE-T1 used in the automotive industry. 
• CR (Copper/Twinax): Operates over short distances of 3 to 7 meters (m) on copper twinaxial cable. This is the industry standard for direct-attach copper (DAC) cables pre-terminated with SFP or QSFP transceivers, commonly used for direct top-of-rack (ToR) switch-to-server links in data centers.
• SR (short reach): Uses multimode fiber at the 850 nm wavelength for distances ranging from 60 to 550 m, depending on the speed and multimode type (OM3 vs. OM4). Examples include 10GBASE-SR for 10 Gb/s operation (300 m on OM3 and 400 m on OM4) and 25GBASE-SR for 25 Gb/s operation (70 m on OM3 and 100 m on OM4). To achieve higher speeds, SR applications leverage parallel optics technology, in which signals are transmitted over multiple fibers.
• VR (very short reach): Optimized for shorter lengths of multimode distances at the 850 nm wavelength, typically 30 m for OM3 and 50 m for OM4. Developed to support 100 Gb/s speeds for intra-rack or adjacent-rack connections in the data center, VR transceivers maximize efficiency by slashing costs and power consumption. VR applications also leverage parallel optics technology to achieve higher speeds.
• DR (data center reach): Transmits up to 500 m over singlemode fiber at the 1310 nm wavelength. Often called “short-reach singlemode,” DR applications use cost-effective, low-power lasers to achieve distances beyond multimode reach. DR applications also leverage parallel optics technology to achieve higher speeds.
• FR (functional/far reach): Reaches up to 2,000 m over singlemode fiber. Designed as a budget-friendly intermediate option, FR applications avoid expensive, high-power long-haul lasers. To achieve higher speeds, FR applications use wavelength-division multiplexing (WDM), in which signals are transmitted over multiple wavelengths on a single fiber.
• LR (long reach): Reaches up to 10 kilometers (km) on singlemode fiber, making it ideal for high-speed LAN, MAN, and campus backbones, as well as for linking geographically dispersed facilities. LR applications also use WDM technology to achieve higher speeds.
• ER (extended reach): Extends up to 40 km on singlemode fiber using high-power lasers. ER applications often serve as the backbone of wide-ranging service provider networks that connect cities and regions. ER applications also use WDM technology to achieve higher speeds.
• ZR (zero-dispersion range): An ultra-long-haul singlemode application reaching at least 80 km, with some iterations stretching to 120 km. ZR applications are reserved for international backbone network connections, including sub-sea or cross-country links. ZR applications use dense WDM (DWDM) technology to achieve higher speeds.

The IEEE also develops Ethernet standards for internal connections used inside networking equipment. KR indicates a copper electrical backplane application, providing ultra-short links of 1 m or less between plugged-in cards within the same switch or router chassis. AUI (Attachment Unit Interface) is an electrical interface between two semiconductor chips on the same printed circuit board (PCB), or between a host chip and a transceiver module. AUI chip-to-chip connections are foundational in modern AI equipment, allowing different processors on the same PCB to communicate. 

Note: While IEEE standardizes these naming conventions for Ethernet, you may encounter non-IEEE suffixes used for proprietary, vendor- and MSA-specific applications. 

Counting Lanes and Wavelengths

For many fiber Ethernet applications, IEEE adds a number to the end of the acronym that indicates how many transmit lanes are used to achieve the total aggregate speed. Because Ethernet technology has rapidly evolved, these numbers are not always consistent. For example, single-lane applications often omit the trailing number entirely, while some explicitly use a “1.” In some contexts, the numbers are replaced with an “X” to signify any lane cont. However, these numbers can still tell us a lot about an application. 

In parallel-optics applications (SR, VR, and DR), the trailing number indicates the number of physical transmit channels. Because every link requires both transmit (TX) and receive (RX) lanes, you can double this number to determine your total required fiber count:  

• 400GBASE-SR4: Uses 4 transmit lanes and requires 8 fibers total (4 TX / 4 RX)
• 400GBASE-SR8: Uses 8 transmit lanes and requires 16 fibers total (8 TX / 8 RX)
• 400GBASE-SR16: Uses 16 transmit lanes and requires 32 fibers (16 TX / 16 RX)

WDM applications (FR, ER, LR, and ZR) collapse traffic onto just two physical fibers by using different wavelengths. For these standards, the trailing number indicates how many operating wavelengths are packed onto each fiber.

• 400GBASE-ER8: Uses 8 distinct wavelengths on each fiber (8 transmitting on one, 8 receiving on the other).
• 200GBASE-ER4: Uses 4 distinct wavelengths on each fiber (4 transmitting on one, 4 receiving on the other).

Ever wonder why we have multiple variants like SR4, SR8, and SR16 running at the same speed? This all comes down to breakthroughs in signaling technology, which have allowed individual lane speeds to climb from 10 Gb/s to 25 Gb/s, 50 Gb/s, and 100 Gb/s. For example, the older 400GBASE-SR16 application relies on 16 separate lanes transmitting at 25 Gb/s each. In contrast, the newer 400GBASE-SR4 application achieves that same 400 Gb/s throughput using just 4 lanes transmitting at 100 Gb/s each. 

Because of this naming convention, you can use the trailing number to figure out the lane speed. If you know the application speed is 200 Gig and the trailing number is 4, you can easily deduce that the application uses 50 Gb/s per lane. You can learn more about how advances in signaling technology have increased lane speeds in our previous blog on the Evolution of High-Speed Fiber Optics.

Note that the IEEE also sometimes uses decimals to denote hybrid architectures that combine parallel optics and WDM. Take 400GBASE-SR4.2 for example. The “SR4” indicates short-reach multimode over 4 physical channels, while the “.2” indicates 2 distinct wavelengths per fiber. By combining parallel optics and WDM technology, 400GBASE-SR4.2 achieves 400 Gig using a 50 Gb/s lane rate over just 8 total fibers (2 wavelengths over 4 TX fibers, 2 wavelengths over 4 RX fibers). 

Choosing Your Best Path Forward

Cracking the IEEE naming code transforms confusing acronyms and numbers into a functional roadmap to decipher an application’s speed, media, distance, and signal type. This helps you choose the right transceivers, fiber type, and connectivity for your specific applications.

The good news is that you don’t have to navigate the landscape alone. Cables Plus USA offers an extensive selection of solutions for Ethernet applications, scaling up to 1.6 Terabit. Whether you require Category copper cables for BASE-T up to 10 Gig, twinax DACs for short-reach CR applications, high-density MTP/MPO connectivity for parallel optics, or duplex cables for singlemode WDM, contact us today for help choosing the right cable and connectivity for your next Ethernet deployment.