Multimode Optical Fiber in Telecommunications

Published by: Research & Development Department, Technologie Optic.ca Inc., April 2026

Introduction

Multimode optical fiber occupies an intermediate position between low-cost premises cabling and high-performance optical communication infrastructure. Its large core diameter, commonly 50/125 µm or 62.5/125 µm, allows multiple guided modes to propagate simultaneously. This physical property makes alignment at connectors and transceiver interfaces less stringent than in single-mode systems, thereby reducing installation complexity and enabling compact, relatively inexpensive optics. For this reason, MMF has been widely adopted in short-reach telecommunications environments such as data centers, enterprise buildings, campus backbones, and certain access-network segments.

The principal limitation of MMF is intermodal dispersion. Because different modes travel along different optical paths and experience different propagation delays, an input pulse broadens as it traverses the fiber. The effect becomes increasingly restrictive as line rate rises, making bandwidth-distance performance the key metric in multimode design. Graded-index core profiles were introduced to mitigate this limitation by reducing differential modal delay, and later generations of laser-optimized multimode fiber were specified to improve performance with 850 nm VCSEL-based transmitters.

From a telecommunications perspective, MMF should not be regarded as a competitor to single-mode fiber for metro or long-haul transport. Rather, it is a strategically important solution for the short-reach portion of optical networks. In this regime, the governing design problem is not ultimate span length but cost-effective delivery of high aggregate throughput over structured cabling infrastructures that must support frequent equipment refresh cycles. Accordingly, the practical value of MMF depends not only on the fiber class itself, but also on the transceiver family paired with it.

Multimode Fiber Fundamentals in Telecom Context

Multimode guidance is typically realized through a graded-index rather than a step-index core. In step-index multimode fibers, higher-order modes travel substantially longer geometric paths than lower-order modes, creating severe pulse broadening. In graded-index fibers, the refractive index decreases gradually from the center of the core toward the cladding, so higher-order modes propagate faster in outer regions. This partially compensates for their longer paths and reduces intermodal dispersion. The graded-index profile is therefore fundamental to modern telecom-grade MMF. As shown in Figure 1, the difference between step-index and graded-index multimode fibers lies not only in their refractive-index distribution, but also in their ability to control modal propagation and reduce pulse spreading.

Comparison of step-index and graded-index multimode fibers showing refractive-index profiles and modal propagation
Figure 1: Comparison of step-index and graded-index multimode fibers, showing their refractive-index profiles and the corresponding effect on modal propagation and intermodal dispersion.

Although long-haul optical transport relies overwhelmingly on single-mode fiber, MMF remains relevant in telecommunications wherever transmission distances are modest and equipment density, upgrade flexibility, and cost are dominant constraints. Examples include in-building distribution systems, multi-tenant premises networks, data-hall interconnects, campus aggregation, and short-reach access segments. In such cases, MMF enables practical reuse of structured cabling while supporting successive generations of Ethernet optics.

The historical evolution of MMF optics follows the transition from LED-based links to laser-based short-reach systems. Early multimode deployments were associated with overfilled-launch LEDs and modest bandwidth. Modern systems rely primarily on 850 nm VCSEL transmitters, which offer low power consumption, high modulation speed, and good manufacturability. As data rates increased further, parallel optics, bidirectional optics, and shortwave wavelength multiplexing were developed to increase aggregate throughput without proportionally increasing fiber count.

OM1 to OM5: Fiber Classes and Physical Characteristics

The OM nomenclature designates standardized multimode fiber classes used in structured cabling and premises optical networks. These categories should be interpreted as minimum performance classes rather than fixed reach guarantees, because the practical transmission distance of a given link also depends on the transceiver type, launch condition, connector and splice loss, and the overall channel budget. As shown in Figure 2, the progression from OM1 to OM5 reflects both structural continuity and performance evolution: OM1 retains the legacy 62.5/125 µm format, whereas OM2 through OM5 use 50/125 µm fibers; OM3 and OM4 are laser-optimized for 850 nm VCSEL-based short-reach transmission, while OM5 extends multimode capability into the wideband 850–950 nm region for wavelength-division applications. Conventional jacket identification also follows this classification, with OM1 and OM2 typically orange, OM3 and OM4 aqua, and OM5 lime green.

Schematic comparison of multimode fiber classes OM1 through OM5 showing core size and jacket color identification
Figure 2: Schematic comparison of multimode fiber classes OM1–OM5, showing nominal core size, typical jacket color identification, and their general evolution from legacy multimode fiber to laser-optimized and wideband multimode fiber.

OM1

OM1 uses a 62.5 µm core and is generally identified by an orange jacket. It is associated with legacy LED-era systems and low modal bandwidth compared with later classes. OM1 can still support some Gigabit Ethernet and short 10 Gigabit links, but its applicability in modern high-speed systems is limited. In present-day telecom practice, OM1 is primarily encountered in legacy building infrastructure rather than new installations.

OM2

OM2 uses a 50 µm core, typically with an orange jacket. It offers improved bandwidth relative to OM1 and supports longer 1G and 10G links, but it remains a legacy class when compared with OM3 and later fibers. OM2 may still be reused where existing infrastructure must be preserved, especially for modest line rates and short distances.

OM3

OM3 is the first laser-optimized multimode class and is typically identified by an aqua jacket. It was designed for efficient operation with 850 nm VCSEL sources and became the practical baseline for 10G, 40G, and early 100G multimode systems. OM3 remains widely deployed in data centers because it offers a strong balance between performance, cost, and compatibility.

OM4

OM4 is also a 50 µm laser-optimized fiber, usually aqua, though some vendors use violet to distinguish it from OM3. OM4 improves effective modal bandwidth and extends the reach of short-reach optics without changing the basic operating wavelength regime. It is commonly selected where longer short-reach links, denser interconnects, or greater upgrade headroom are required.

OM5

OM5, normally lime green, extends the multimode concept into the wideband regime. It preserves OM4-class performance at 850 nm but additionally specifies multimode behavior across a wider shortwave spectrum, enabling multiwavelength transceivers such as SWDM and some bidirectional/WDM-based designs. OM5 is most meaningful when the optics actually exploit wavelengths beyond 850 nm; for ordinary single-wavelength 850 nm SR optics, OM5 does not automatically provide a significant reach advantage over OM4.

Bandwidth, Launch Conditions, and Reach

Bandwidth in multimode systems is generally expressed as a bandwidth-length product, reflecting the fact that usable transmission bandwidth declines as fiber length increases. Two distinct concepts matter. Overfilled-launch bandwidth reflects legacy LED-type excitation, while effective modal bandwidth better predicts the performance of modern laser-based systems. This distinction is essential because the migration from OM1 and OM2 to OM3, OM4, and OM5 is not simply a change in core size or color coding; it is a change in how the fiber is engineered to support high-speed optical launches.

In engineering practice, reach is governed simultaneously by attenuation and dispersion. Consequently, a transceiver cannot be matched to a fiber class solely by nominal data rate. The modal bandwidth of the installed fiber, the launch wavelength, the connector count, and sometimes the need for mode-conditioning patch cords all affect the reliable operating distance.

Multimode Transceiver Families and Maximum Reach

From a telecom engineering perspective, multimode transceivers are most clearly understood when grouped by line rate rather than by vendor-specific variations. To keep the discussion consistent, only the principal MMF families are considered here: 1000BASE-SX at 1G, 10GBASE-SR and 10GBASE-LRM at 10G, 25GBASE-SR at 25G, 40GBASE-SR4 and 40G BiDi at 40G, 100GBASE-SR4 and 100G BiDi at 100G, and 400GBASE-SR8 and 400GBASE-SR4.2 at 400G. These interfaces represent the main practical milestones in the evolution of multimode telecom.

At 1 Gb/s, the reference multimode optic is 1000BASE-SX, operating at 850 nm over duplex MMF. It typically supports approximately 275 m on OM1 and 550 m on OM2, while longer reach is possible on higher-bandwidth 50 µm fibers. At 10 Gb/s, 10GBASE-SR became the standard short-reach MMF interface, offering about 33 m on OM1, 82 m on OM2, 300 m on OM3, and 400 m on OM4 or OM5. In parallel, 10GBASE-LRM was developed to support 10 Gb/s transmission over installed legacy MMF, typically up to 220 m, with mode-conditioning often required on OM1 and OM2.

At 25 Gb/s, 25GBASE-SR extended the duplex 850 nm VCSEL model and typically supports 70 m on OM3 and 100 m on OM4 or OM5. For 40 Gb/s, the dominant standardized MMF interface is 40GBASE-SR4, which uses parallel optics and usually reaches 30 m on OM2, 100 m on OM3, and 150 m on OM4 or OM5. A second important option is 40G BiDi, which operates over duplex LC multimode cabling and is mainly intended to preserve existing duplex MMF plants, typically reaching 100 m on OM3 and 150 m on OM4 or OM5.

At 100 Gb/s, the principal multimode interface is 100GBASE-SR4, which supports roughly 70 m on OM3 and 100 m on OM4 or OM5. Duplex-fiber migration solutions also exist in the form of 100G BiDi-type optics, which generally provide similar reach on OM3 and OM4, with some implementations extending up to 150 m on OM5. At 400 Gb/s, multimode solutions split into two major categories: 400GBASE-SR8, which uses parallel multimode lanes and typically reaches 70 m on OM3 and 100 m on OM4 or OM5, and 400GBASE-SR4.2, which combines multimode transmission with wavelength-division principles and can extend to about 150 m on OM5. Among currently deployed multimode solutions, SR4.2 is one of the clearest examples of an optic that can make practical use of OM5 wideband capability.

Comparative Table of MMF Transceiver Reach Across OM1–OM5

Table 1 compares the principal multimode transceiver families by data rate and typical maximum reach over OM1 to OM5 cabling. These values should be interpreted as practical design guidance, since actual performance also depends on insertion loss, connector count, and vendor implementation.

Table 1: Typical maximum reach of major multimode transceiver families across OM1–OM5 fiber classes
Speed Transceiver / Interface OM1 OM2 OM3 OM4 OM5
1G 1000BASE-SX 275 m 550 m up to 1 km up to 1 km up to 1 km
10G 10GBASE-SR 33 m 82 m 300 m 400 m 400 m
10G 10GBASE-LRM 220 m 220 m 220 m 220 m 220 m
25G 25GBASE-SR n/s n/s 70 m 100 m 100 m
40G 40GBASE-SR4 n/s 30 m 100 m 150 m 150 m
40G 40G BiDi n/s n/s 100 m 150 m 150 m
100G 100GBASE-SR4 n/s n/s 70 m 100 m 100 m
100G 100G BiDi n/s n/s 70 m 100 m up to 150 m
400G 400GBASE-SR8 n/s n/s 70 m 100 m 100 m
400G 400GBASE-SR4.2 n/s n/s 70 m 100 m 150 m

Abbreviation: n/s = not standardized or not commonly specified for that OM class in normal deployment guidance.

Connectors, Cabling Architecture, and Practical Deployment

The transceiver family largely determines the connector format used in MMF telecom. Duplex serial interfaces such as SX, SR, LRM, LX4, and many BiDi modules usually employ LC connectors. Older systems may use SC or ST. Parallel interfaces such as SR4 and SR8 typically require MPO/MTP connectors and pre-terminated multifiber trunks.

This distinction has major operational consequences. Duplex LC infrastructures are easier to maintain and are common in legacy data halls and enterprise backbones. MPO infrastructures enable higher aggregate bandwidth and lane breakout, but they require stricter polarity management, cleaner installation practice, and different patching logic. The choice between duplex and parallel MMF is therefore not only optical but architectural.

Legacy reuse also requires attention to mode-conditioning rules. Long-wavelength laser interfaces operating over older MMF, such as 1000BASE-LX/LH, 10GBASE-LX4, and 10GBASE-LRM, may require mode-conditioning patch cords on FDDI-grade, OM1, and OM2 fibers. By contrast, these conditioning cords should not be used with OM3 or later laser-optimized plants.

Selection Guidance Across OM1–OM5

A rational selection strategy begins with the installed fiber class. If the site is built on OM1 or OM2, the designer is operating in a legacy-constrained regime. In this case, 1G SX, short 10G SR, or 10G LRM are realistic options, while 40G and 100G operation is generally impractical except in highly specialized or vendor-specific cases.

If the site uses OM3, the network gains access to mainstream VCSEL-era optics: 10G SR to 300 m, 25G SR to 70 m, 40G SR4 to 100 m, 40G CSR4 to 300 m, and 100G SR4 to 70 m. OM3 therefore remains viable for many contemporary telecom spaces, especially when distances are modest.

OM4 extends these margins and is often the preferred choice when 10G-to-100G migration is expected. It supports 10G SR to 400 m, 25G SR to 100 m, 40G SR4 to 150 m, 40G CSR4 to 400 m, and 100G SR4 to 100 m, while also enabling better performance with certain BiDi and high-order multimode optics.

OM5 should be selected when the design explicitly anticipates wideband or multiwavelength MMF optics, such as SWDM or 400G SR4.2-type operation, or when duplex reuse at higher rates is strategically important. If the application uses only conventional 850 nm SR transceivers, OM5 does not necessarily provide a meaningful advantage over OM4.

Conclusion

Multimode fiber continues to play a significant role in telecommunications wherever short reach, structured cabling compatibility, and cost-efficient optics are central design goals. The technical evolution from OM1 and OM2 to OM3, OM4, and OM5 mirrors the broader progression from LED-based links to VCSEL-driven short-reach architectures, then onward to parallel optics, BiDi reuse strategies, and shortwave wavelength multiplexing.

The practical meaning of MMF in telecom cannot be understood from fiber class alone. OM1 through OM5 define the transmission medium, but the usable system performance emerges only when that medium is paired with an appropriate transceiver family. For legacy plants, the key choices are often 1000BASE-SX, 10GBASE-SR, and 10GBASE-LRM. For modern structured cabling, the dominant families include 25G SR, 40G SR4 and BiDi, 100G SR4 and BiDi derivatives, and newer 400G multimode interfaces. In all cases, maximum range is governed jointly by wavelength, launch type, modal bandwidth, connector architecture, and channel loss.

From an engineering standpoint, OM3 and OM4 remain the practical baseline for most modern MMF telecom deployments, while OM5 is best justified when a true wideband multimode roadmap is intended. The enduring value of multimode technology lies in this combination of pragmatic optics, scalable cabling, and strong compatibility with dense short-reach communication environments.

Mohammad Bakhtbidar, PhD
Head of the Research & Development Department
Technologie Optic.ca Inc.