The battle between the fibres

There is a huge difference between multimode (MM) and singlemode (SM) fibres; MM fibres have a much larger core – typically 50 micrometres – while SM fibres are much smaller, with the core measuring nine micrometres. MM fibres are used over shorter distances, for example, within a building or on a campus. Its main disadvantage is its modal dispersion, which means the optical signal travels a different distance depending upon the mode it follows within the fibre. This is not the case with SM fibres, which are used to cover much longer distances. As a result, the transmission distance for MM fibres is much lower than for SM fibres.

Within a data centre, it’s not just about bandwidth and distance – it’s also about the total cost of deployment. SM fibres require more expensive laser sources, while MM fibres can use lower-cost light sources like LEDs (Light Emitting Diodes) and VCSELs (Vertical Cavity Surface Emitting Lasers). The connectors can also differ in cost. MM fibres allow an alignment tolerance of 3-5 micrometres, while SM fibres require a connector with 0.5-1 micrometre tolerance. In a connector, or actually, in general, lower tolerance means higher cost of components!

For MM fibres we have seen a transition in types of MM fibre. Essentially, a 50 micrometre core has superseded the previously dominant 62.5 micrometre core. Both typically have had an orange or grey cable jacket. MM OM2 fibres have a bandwidth-distance product of 500 MHz*km, while MM OM3 transmits 2000 MHz*km and MM OM4 transmits 4700 MHz*km. These bandwidths only apply for a wavelength of 850nm which is mainly used in laser-based applications (VCSELs). OM3 and OM4 cable jackets generally have an aqua colour.

The next step in MM fibres could be Wide Band MM Fibre (WBMMF). WBMMF currently available on the market can support four wavelengths. In combination with Wavelength Division Multiplexing (WDM), it expands the capacity of a MM fibre by a factor of four, and maybe more in the near future. This means higher bandwidth density and less fibres. The WDM functionality could be integrated into the transceiver and, as such, be invisible to the installers. The question yet again is, what is the total cost of deployment of WBMMF? When compared to MM OM4: connector costs can be similar, fibre costs can be slightly higher – but hopefully not by much – and transceiver costs will be more, due to the integration of WDM functionality. As an enabled WBMMF channel replaces four traditional channels, to be commercially viable, the cost will need to be a factor of four or less.  

For SM fibre (typically a yellow jacket), several developments are happening. Extensive research is being done into the multicore SM fibre, which today, contains four or eight separate cores within the same 125 micrometre outer diameter fibre. An improvement in each core would mean a bandwidth increase within a fibre of a factor of four or eight. Again, the main question here is, what is the total cost of deployment? When compared to MM fibre: connector costs will be higher because of its required rotational sensitivity, fibre costs will be higher due to the more complex preform, and transceiver costs will be higher, also due to the accurate alignment requirement for multiple cores. The question is, again, is it a factor of four, eight, or less?

Will there be a clear winner between MM and SM fibre? I don’t think so. Both of these solutions will find their way into our data centres, but it will depend on specific configurations helping us to gain density. More about density next time! Hope to ‘meet’ you again!