University and hospital campuses that installed OM3 multimode fibre backbones a decade ago are now hitting optical power budgets that 100G transceivers simply won't forgive. The upgrade path isn't a single answer — it depends on distance, existing duct capacity and how far ahead the campus needs to plan.
Why OM3 Runs Out of Headroom at 100G
OM3's modal bandwidth specification was set with 10G and early 40G applications in mind, and its optical power budget shrinks fast as speed increases. Running genuine 100G over OM3 typically requires parallel-fibre transceivers using multiple fibre pairs simultaneously, over distances far shorter than the same equipment achieves on OM4 or singlemode OS2 — on a campus-scale backbone spanning several hundred metres between buildings, OM3's reach at 100G frequently falls short of what the physical layout actually needs, forcing costly intermediate regeneration points that weren't part of the original design.
OM4 vs OS2: The Real Decision Point
- OM4 multimode extends reach and power budget over OM3 at 100G, and remains cheaper on the transceiver side (multimode optics cost less than singlemode), making it a reasonable middle-ground upgrade for shorter campus links.
- OS2 singlemode supports much longer distances and a clear upgrade path to 400G and beyond without a further cable replacement — the fibre itself is a modest cost premium over multimode, though singlemode transceivers currently carry a higher unit cost than multimode equivalents.
- For any new campus backbone install or full replacement, OS2 is the more future-proof default — the cost difference is in the transceivers, which get cheaper over the technology's life, not in the fibre, which is expensive and disruptive to replace again.
- BiDi (bidirectional) transceivers send and receive on a single fibre strand using different wavelengths, effectively halving fibre count needed for a given link — valuable where existing duct or conduit capacity constrains how many additional strands can be pulled.
Design takeaway: Where budget allows a full backbone replacement rather than an incremental upgrade, specify OS2 singlemode — it defers the next fibre replacement cycle by a decade or more, and the fibre itself is the expensive, disruptive part of any campus backbone project, not the transceivers.
Before Committing to a Backbone Refresh: Audit the Physical Pathway
The most consequential step before any Australian campus commits capital to a backbone upgrade is a pit-and-pipe audit — confirming actual available duct capacity, condition of existing conduit, and pit access along the intended route. It's common on campuses with decades of incremental infrastructure additions to discover that documented duct capacity doesn't match what's actually usable, which can change the entire cost and staging plan for a backbone refresh before a single fibre metre is ordered.
Frequently Asked Questions
Why does OM3 multimode fibre struggle to support 100G links?
OM3's modal bandwidth and reach specifications were set for 10G and early 40G applications. Running 100G over OM3 requires parallel-fibre transceivers (using multiple fibre pairs) over much shorter distances than the same link achieves on OM4 or singlemode OS2, which quickly becomes impractical on campus-scale backbone distances.
Should an Australian campus upgrade to OM4 or go straight to OS2 singlemode?
For new campus backbone runs, OS2 singlemode is the more future-proof choice — it supports much longer distances and higher future speeds (400G and beyond) without a further cable replacement, at a modest cost premium over multimode for the fibre itself, though singlemode transceivers currently cost more than multimode equivalents.
What is BiDi optics and why does it matter for a backbone upgrade?
Bidirectional (BiDi) transceivers send and receive on a single fibre strand using different wavelengths, rather than needing a separate strand for each direction — halving the fibre count needed for a given link, which can be valuable when upgrading a campus backbone constrained by existing duct or conduit capacity.