01 / The caseThe layer everything else depends on
Structured cabling is a standards-based, hierarchical system of cables, connectors and pathways, designed once and as a whole, so that any device can be connected to any other through a predictable patching scheme.
The opposite is point-to-point sprawl: a new run for every need, no documentation, no headroom. The difference does not show on day one. It shows on every change, every fault and every upgrade for the next ten years.
It outlives the kit
Servers and switches are refreshed every three to five years; the cabling and pathways behind them often last fifteen to twenty-five. You are cabling for generations of hardware you have not bought yet.
It is where faults hide
A large share of network problems are physical: a marginal connector, a kinked fibre, an untested run, the wrong patch lead. They are intermittent, hard to find, and look like “the network being slow”.
Cheap to lay, costly to redo
Cabling is a small fraction of build cost. Ripping it out of a live hall, or working around a mess for years, is not. The economics favour doing it properly, once.
It sets your speed ceiling
The media you install fixes how fast and how far you can run. Install the wrong fibre or copper and the upgrade you want in year four needs a re-cable, not a transceiver swap.
02 / The anatomyInside a structured cabling system
The standards, TIA-942 in North America and ISO/IEC 11801-5 internationally, define a hierarchy of spaces and the cabling that links them. Traffic flows from the carriers at the edge of the building, through distribution areas, to the equipment in the cabinet. Two cabling types do the work: backbone, between distribution areas and almost always fibre, and horizontal, the final run to the equipment.
Main distribution area. The core of the data centre: the main cross-connect, core routers and spine switches. Every backbone run starts here.
Backbone and horizontal
Backbone links the distribution areas and carries aggregated traffic: high-count fibre, sized for the next speed. Horizontal is the last hop, where copper or fibre meets the server.
Patch panels and the cross-connect
Permanent cabling terminates on patch panels; short patch leads make the actual connections. Moves and changes happen on the patching, never by re-pulling permanent cable. That is the point of “structured”.
Pathways and containment
Tray, basket, ladder and underfloor or overhead routes keep cable supported, separated from power and accessible. Bad pathways are where bend-radius and congestion problems start.
03 / The mediaCopper, fibre and how far they reach
Three media do almost all the work in a modern hall: balanced copper (twisted pair, RJ45), twinax DAC (direct-attach copper, for in-rack links) and optical fibre, multimode for short reach and single-mode for everything longer. The right choice trades distance, speed, power and cost.
| Media | Typical use | Reach | Notes |
|---|---|---|---|
| Cat6A copper | Horizontal to the server, 10GBASE-T | 100 m | Cheap, RJ45, no transceiver, but power-hungry and in practice capped at 10G |
| Cat8 copper | Top of rack, 25/40GBASE-T | ~30 m | Short reach only; largely superseded by DAC inside the rack |
| Twinax DAC | Server to ToR switch, in-rack | 1–7 m | Lowest cost and power for short hops; passive or active |
| OM4 multimode | Intra-row, short backbone | ~100 m at 100G+ | Cost-effective VCSEL optics; reach shrinks as speed rises |
| OM5 multimode | Short reach, SWDM | ~100–150 m | Wideband multimode for shortwave WDM; a modest gain over OM4 |
| OS2 single-mode | Backbone, inter-room, campus | 500 m–10 km+ | The future-proof choice: the fibre rarely limits the speed |
The headline rule: multimode is cheaper per link but limited, and its reach shrinks as speeds climb; single-mode costs a little more in optics but is effectively speed-proof. Many operators now default to single-mode for the backbone so that the cabling is never the reason they cannot upgrade.
Pick a link speed, then set the distance
The four media light up where they can carry that speed over that distance, and drop out where they cannot.
Reaches are the standardised maxima for each Ethernet PMD and cabling class, rounded. Real links depend on insertion loss, connectors and installation quality.
04 / TopologiesHow the racks are cabled
Where you put the switches decides how much cable you pull, how you patch and how you cool. The three classic patterns each trade cabling volume against flexibility and switch count.
ToR
Top of rack
A switch, or a pair, in every cabinet; servers connect with short DAC. Minimal copper and simple per-rack containment, but more switches to manage and more uplink fibre. The default for dense, modern halls.
EoR
End of row
Aggregation switches at the end of the row, with every server homed back to them. Fewer, larger switches, but long horizontal runs and heavy cable bundles down the row.
MoR
Middle of row
The end-of-row idea with the switch mid-row, halving the worst-case cable length. A practical compromise between cable volume and switch count.
AI and high-density compute have pushed the industry firmly towards top of rack with a single-mode backbone: keep the high-speed copper hops tiny, server to ToR over DAC, and run fibre, a lot of it, back to the spine. Which is where the speed roadmap gets interesting.
05 / The roadmapFrom 10G to 1.6T
Ethernet speed has climbed roughly tenfold a decade, and the way it gets there has changed. Early jumps added lanes: four 10G lanes made 40G. Modern jumps come from faster lanes: NRZ gave way to PAM4, two bits per symbol, lane rates went from 10 to 25, 50, 100 and 200G, and reach over copper and multimode shrank at every step.
Reach on a logarithmic scale. Figures are the standardised PMD reaches, rounded for clarity; standards evolve.
Notice the pattern as you climb: copper drops out above 40G, multimode reach falls towards the length of a single row, and single-mode carries on. That is the structural reason high-speed halls are going single-mode: the fibre you lay today still works at the next two speeds.
06 / What comes nextFaster, denser, closer to the chip
The drivers are AI clusters and the heavy east-west traffic between GPUs. 800G is shipping in volume, 1.6T (IEEE 802.3dj) is being standardised and 3.2T is on the roadmap. You cannot just keep adding lanes; power and density force the optics themselves to change. These are the shifts a plant laid today should anticipate.
- 01
200G per lane
The step behind 1.6T: ~200G lanes on 224G SerDes with PAM4. Fewer fibres for the same bandwidth, but tighter loss budgets, so clean, well-tested cabling matters more.
- 02
Linear-drive optics
Removing the DSP from the transceiver to cut power and latency, significant for AI fabrics where tens of thousands of links make every watt count.
- 03
Co-packaged optics
Optics on the same package as the switch ASIC, replacing power-hungry electrical traces. It brings fibre right up to the silicon and changes how the front of the rack is cabled.
- 04
Hollow-core fibre
Light travels about 50% faster through air than glass. Hollow-core cuts latency and is moving from the lab into early deployment, led by latency-sensitive AI and trading workloads.
- 05
Single-mode everywhere
As multimode reach shrinks with each speed, the industry is standardising on parallel single-mode (DR) inside the hall: the most future-proof choice at install time.
- 06
Higher-count, pre-terminated
Base-8 and base-16 MPO trunks, pre-tested and ready to plug in, are replacing field termination: faster to deploy, denser and cleaner to certify at AI-fabric scale.
AI and GPU halls push all of this further: 800G to 1.6T fabrics, racks drawing 40 to 130 kW, and cable volumes that make containment and airflow a first-order design problem. Planning an AI/GPU-ready data hall goes deep on 800GbE networking and the physical design that supports it.
07 / Doing it rightWhat separates a good install from a liability
The media choice is the easy part. What makes a cabling plant easy to operate, or a decade-long headache, is the discipline around it. These are the things that matter most when you assess an existing install.
Labelling and certification sit at the top for a reason: an unlabelled, untested plant is one where every future change is a guess. A proper install is certified, with every link tested to standard and the result saved, and documented, so the person who works on it in five years is not reverse-engineering your decisions. Certification is also what makes a manufacturer’s long-term system warranty available: no certified results, no warranty.
08 / How Optronix helpsDesigned right, installed clean, certified and warranted
Structured cabling is the physical foundation under everything else we do, and it starts with design. We design, install, test, certify and warrant cabling systems to TIA and ISO standards, across Europe, North America, the Middle East and APAC: from a single row expansion to a greenfield hall, or untangling a plant that grew without a plan.
Designed for where you are heading
Our team has designed structured cabling systems for some of the largest hyperscale data centres in the world, where a single design has to scale to hundreds of thousands of links and survive several speed generations. The same discipline goes into every project, whatever its size.
A 25-year manufacturer system warranty
Optronix is a certified installation partner of the major cabling manufacturers, so the systems we design and install can carry their 25-year manufacturer system warranty, covering components, performance and applications assurance. That cover is only available through accredited installers, on properly installed and certified plant.
- 01
Design and specify
Media, topology, counts and pathways sized for the speeds you will run, with headroom to upgrade on a transceiver swap.
- 02
Install and terminate
Copper and fibre, pre-terminated trunks, patch fields and containment, installed cleanly and to standard.
- 03
Test and certify
Every link tested and certified on TREND Networks LanTEK IV-S testers, with the results documented.
- 04
Document and manage
Labelling, as-built records and DCIM-ready documentation, plus moves, adds and changes as part of operations.
Related services
Structured cabling
Tell us what you run now and where you want to get to.
We will design and install a cabling plant that gets you there: tested, certified, documented and backed by a 25-year manufacturer warranty.
Sources and notes
- TIA-942 (Telecommunications Infrastructure Standard for Data Centers) and ISO/IEC 11801-5: the structured-cabling hierarchy, spaces and media classes.
- IEEE 802.3 Ethernet standards, including 802.3df (800GbE, 2024) and 802.3dj (1.6TbE, in development); media reaches are the standardised PMD figures.
- Ethernet Alliance Technology Roadmap: the speed timeline and the path to 1.6T and 3.2T and 200G-per-lane signalling.
- General industry practice for topologies (ToR, EoR, MoR), DAC and optics selection, and emerging technologies (LPO, CPO, hollow-core fibre).
General guidance, not a design specification. Reaches are the standardised maxima for the relevant Ethernet PMD and cabling class; achievable distance on a real link depends on insertion loss, connectors and installation quality, and standards evolve. Confirm figures against the relevant standard and a tested channel.