Blog

Hengtong Customised Optical Cable Solutions for Australian Data Centres

I'm an optical cabling engineer with more than ten years' experience in designing and delivering fibre systems for data centres and carrier networks, including several projects in the Australian market.

In April 2025 we received an enquiry from a major data centre operator in the Sydney region. They were planning a new phase of construction and expansion and wanted to standardise the entire optical cabling system across a multi–data hall campus – from outdoor campus fibre, through in-building backbone, right down to high-density MPO/MTP cabling inside the halls. I was appointed as technical lead, responsible for solution design, product selection, delivery and technical support.

This article is a technical case study of that project and a summary of how we approached it as engineers. I'll focus on how Hengtong built an end-to-end, customised optical cable solution for an Australian data centre campus, based on outdoor fibre, indoor fibre and MPO/MTP high-density cabling.

An optical cabling system is never designed in isolation. It has to match the business model and the way the site will be operated. From the projects I've worked on, most Australian data centres fall into three broad roles – many campuses are a mix of all three

Colocation / carrier-neutral data centres

These are a major part of the Australian market. The operator provides space, power, cooling and a basic network environment to multiple tenants, such as:

Public and hybrid cloud providers

Video and content platforms

Banks and other financial institutions

Large enterprises and SaaS providers

The cabling here has to be flexible enough to handle many tenants, but clearly segregated for security and billing. That drives up requirements for fibre capacity, redundancy and structured cabling.

Hyperscale / cloud data centres

Built by cloud or internet companies for their own platforms, these sites typically host:

Large compute and storage clusters

Core cloud infrastructure

Big data, AI, CDN and other high-bandwidth workloads

They are very large with extremely high port density. Networks are usually based on spine–leaf or other Clos-style designs. From a cabling point of view, the requirement is simple:

Run 100G/400G now and provide a natural path to 800G and beyond.

Edge / regional data centres

These are closer to end users or where data is generated, to reduce latency and meet local data rules. They're smaller than big colos or hyperscales but just as demanding on reliability and fibre path redundancy – sometimes more so, because on-site support in regional areas is expensive and slower to respond.

The Sydney campus in this case study is primarily a large colo site, but with cloud and content tenants as well. In practice, all three types of requirements show up at the cabling layer at the same time.

Even with different business models, the core expectations for cabling in Australian data centres are similar.

A multi–data hall campus in Sydney

This project was set in a large, multi–data hall data centre campus in the Sydney region:

Multiple data centre buildings, some in production, others under construction or reserved for future phases

Each building split into several independent data halls, planned by tenant type, power density and security level

A diverse tenant mix: international cloud providers, local financial institutions, content platforms and large enterprises

One cabling architecture for the whole campus

From a cabling perspective the key point is:

It's not "one room, one design" – the cabling has to be planned for the whole campus lifecycle.

The customer's goals were clear:

Use this expansion phase to create a single, coherent cabling architecture – from outdoor carrier / campus backbone fibre through in-building backbone, down to MPO/MTP cabling and in-rack patching.

Make sure that new halls, new tenants or upgrades from 100G/400G to 800G don't require ripping out the backbone.

Treating the campus as a single end-to-end optical path

So we treated the entire path as a single end-to-end optical link:

Outdoor carrier / campus OSP fibre
→ Campus and inter-building backbone
→ Building entrances and outdoor–indoor transition
→ Indoor LSZH backbone (risers and corridors)
→ MPO/MTP structured cabling inside data halls
→ In-rack patching and port management

Hengtong's role was to provide an end-to-end solution along this chain – from products to design, pre-termination and testing support – rather than just "selling fibre by the kilometre".

 

Outdoor–indoor continuity

Outdoor OSP cables need mechanical strength, moisture resistance and rodent protection; indoor cables prioritise fire performance, LSZH and flexibility. If you simply "hard cut" from one to the other at the building entry, that transition becomes the weakest point and a magnet for faults. We have to think carefully about how the outdoor cable is terminated, whether an indoor/outdoor hybrid is used, and how fire-stopping and cable density are handled around the entry.

Redundancy and capacity planning

The fibre system can't be sized just for "today's core count". We need diverse physical routes, clear allocations for different halls and tenants, and spare capacity for phases two and three. Done properly, later expansions can break out from existing backbones rather than relaying new ones.

High-density connectivity in the halls

Dozens or hundreds of cabinets, each with many fibre ports, plus tight 100G/400G/800G loss budgets – that combination means loose fibres and random patching aren't an option. Factory-terminated MPO/MTP trunks, high-density patch panels and modular cassettes are the only realistic way to keep density and manageability under control.

Compliance and acceptance

Fire ratings, smoke and toxicity, routing rules, fire compartments – all have to align with Australian data centre and building standards. If this isn't thought through in design, it shows up later as painful rework during inspections.

 

Construction and operations constraints

Skilled fibre crews are limited and schedules are tight. After handover, the system is usually maintained by a general operations team, not fibre specialists. That pushes us towards solutions that are easy to install correctly and easy to manage day-to-day.

 

 

 

 

With those challenges in mind, we built an integrated solution with a few simple design principles.

 

End-to-end thinking

We designed the full chain as one system:

Campus OSP routing and fibre counts

Transition, protection and fire treatment at building entries

Indoor LSZH backbone distribution between floors and halls

MPO/MTP structured cabling and in-rack patching

All of this sat on the same link diagrams and loss budgets.

 

Layered, but coordinated

We still divide the system into four layers:

Outdoor OSP: between buildings and to carriers

Entrance facility: outdoor–indoor transition

Indoor backbone: LSZH risers and corridors

In-hall structured cabling: MPO/MTP trunks, cassettes and patching

Each layer has its own focus – mechanical protection, fire performance, density, loss budget – but fibre counts, interfaces and labelling are consistent end-to-end.

 

"Install once, use for years"

Halls and tenants come in phases; the backbone should not. We designed with a 5–10 year window in mind: spare capacity outdoors and indoors, room for extra cables in trays and conduits, and an MPO/MTP architecture that can scale from 100G/400G to 800G without structural changes.

 

Standard building blocks, customised parameters

We use standard, proven components:

OSP cable constructions

LSZH riser and distribution cables

Mature MPO/MTP panels and cassettes

On top of that we customise fibre counts, lengths, pre-termination styles and labelling to match the real campus layout. Standardisation keeps things simple; customisation keeps them efficient.

Before talking about individual products, we worked with the customer on a high-level topology diagram answering one question:

From the carrier entry to a rack port, what path does a fibre take?

Top to bottom, the topology is:

With one diagram, everyone can see how far a link travels, what type of cable and connectivity is used at each layer, and which layers need more capacity when new halls or rows are added.

For this campus, the "foundation" of the whole system is the outdoor backbone between buildings and to the carrier hand-off points. If this layer isn't designed properly, it doesn't matter how tidy the indoor cabling looks; reliability will always be at risk.

We mainly used Hengtong loose-tube, steel-tape armoured outdoor cables – for example:

• Underground Fibre Optic Cable for buried routes
• Duct Fiber Optic Cable for ducts
• Direct Bury Fiber Optic Cable where extra mechanical protection was required
• Anti Rodent Fiber Optic Cable with steel tape or glass yarn on rodent-prone routes

These provide high mechanical strength, good environmental resistance via water-blocking structures, and rodent protection where needed.

Based on the campus plan and the number of halls per building, we specified 96-, 144- and 288-core versions to avoid running multiple small cables in parallel.

For routing we combined:

Dual-path designs between buildings – a primary duct route and a secondary, physically diverse route to create a ring-like, cut-resistant structure

Dual-path links to the carrier meet-me point, so a single fault cannot take down critical services

Fibre groups were reserved for key halls and tenants, with healthy spares in each backbone for future halls, rows, tenant turn-ups and inter-site connections. Slightly oversizing fibre counts up-front is far cheaper than relaying a campus backbone later.

At quality level we used:

Factory tests to IEC 60794 and ITU-T G.652/G.657 for optical and mechanical performance

Full-path OTDR and continuity/labelling checks on site for every backbone

The operations team received route diagrams, fibre allocation tables and OTDR and insertion-loss reports – documents they now use regularly for troubleshooting and expansion.

The outdoor backbone answers "how do we connect building to building", but the outdoor–indoor transition decides whether the link is safe, compliant and maintainable at the point it crosses the wall.

Each building uses a dedicated entrance room / entrance facility where:

• OSP cables are mechanically secured and strain-relieved
• Splices transition them into indoor/outdoor hybrid or LSZH backbone cables
• Space and slack are reserved for future testing, re-splicing and route changes
• The entrance area also falls under the site's security and access controls.

Between the entrance facility and the riser we chose Hengtong indoor/outdoor hybrid cables such as Indoor Outdoor Round Drop Cable. These retain mechanical protection and weather resistance around the entry while using LSZH sheaths so they can be treated as indoor cables in the fire strategy. They are ideal for entrance rooms, weak-current shafts and the bottom of risers.

At fire-rated walls and floors:

• Compliant fire sleeves and sealing materials are used at each penetration
• Bend radius and working space are controlled so future cable work doesn't require breaking everything open
• Cable types, fire ratings and sealing methods are clearly recorded on drawings and in handover documents
• Thinking about these details at the design stage greatly reduced rework during fire and building inspections.

Inside the building, LSZH indoor backbone cables take over, running from the main distribution area out to each hall and row.

We used Hengtong indoor families such as:

Optical Cables for Vertical Wiring in Buildings and Riser Fiber Optic Cable in risers

Indoor Multi Core Tight-buffered Cable and Easy Branches Indoor Riser Cable in corridors and into halls

They provide LSZH sheaths, compact flexible constructions and clear fibre colour codes for large-scale splicing and patching.

From the MDA we designed:

Vertical riser backbones to floors with data halls

Horizontal backbones in trays or floor voids to each HDA/ZDA

Key paths have alternate routes for resilience and future capacity.

Fibre counts were sized per hall, based on rack counts, port density and service types, with spare capacity aligned to 3–5 year growth plans and extra headroom for especially critical areas. That way, later expansions can often break out from existing backbones instead of installing new ones.

During installation we emphasised:

Route control to avoid strong-power trays, high-temperature zones and vibration points

Respecting bend-radius and pulling-tension limits

A consistent labelling scheme (building–floor–hall–cable ID) on cables, closures and panels, matched to allocation tables, using durable labels

These details are what make later MAC work – moves, adds and changes – efficient and low-risk.

The final layer is the part everyone sees: high-density MPO/MTP cabling inside the halls, which determines how fibres present at device ports and how easy the system is to operate.

We used Hengtong MPO/MTP pre-terminated products, including:

Factory-terminated MPO/MTP trunks and MPO jumpers for spine–leaf and row-level interconnects

MPO–LC / MPO–SC / MPO–FC fan-out jumpers to connect to different device interfaces

Trunks were specified in 12-, 24- and 48-core versions, using OS2 or OM4/OM5 fibre to support 100G/400G today and a clear path to 800G. MPO/MTP polarity and termination schemes were standardised to avoid confusion at cassettes and devices. All assemblies were 100% IL/RL-tested at the factory.

Inside the halls we adopted a structured approach:

High-density 1U/2U fibre panels in row distribution or aggregation racks

MPO/MTP cassettes breaking one or more MPO ports into multiple LC/CS ports

Short patch leads from those ports to switches and servers

This keeps trunks fixed in known locations, limits day-to-day work to the cassette and patch-lead layer, and gives a clear layout that new staff can understand quickly.

On performance we built loss budgets per channel, including outdoor and indoor fibre attenuation, splices, patch panels and MPO/MTP and LC/CS connectors. Where paths were longer or had more connections, we used ultra-low-loss MPO/MTP components to keep total loss within comfortable margins for 40G/100G/400G/800G optics. Factory test reports and on-site end-to-end loss and OTDR tests were used to confirm real-world performance.

At the device, we standardised on LC/CS patch leads using G.657.A2 bend-insensitive fibre from Hengtong's Fibre Optic Jumper / Patch Cord range, with different jacket colours for different functions and standard lengths (1 m, 1.5 m, 2 m) to avoid excess slack.

The net result is a hall-level cabling system that can carry high-speed services and still remain visible, manageable and easy to expand.

Products and designs are only part of the story. Delivery depends on how well they are installed and accepted.

For this project we combined remote and on-site support:

• Pre-construction briefings with contractors and fibre crews on overall architecture, installation rules and MPO/MTP handling
• On-site inspections and demonstrations of splicing, termination and fire-stopping on critical routes, entrance rooms, risers and halls
• Support for OTDR and end-to-end loss testing, plus compilation of as-built drawings, test reports, allocation tables and labelling rules

From my perspective as an engineer, success is when the system continues to run stably under the customer's own team after we leave. Good support and documentation are what make that possible.

Looking back on a project like this, I mainly ask three questions: Does it perform optically? Is it mechanically and environmentally robust? And does it meet standards and local compliance?

On optical performance, design-stage loss budgets for 100G/400G (with room for 800G) were confirmed by acceptance tests; key backbones and 400G-ready channels came in comfortably within budget. Spot checks after handover showed stable loss across seasonal and operational changes.

On mechanical and environmental reliability, the combination of armoured, water-blocked OSP cables and LSZH indoor backbones, together with controlled routes and bend radii, gives the system enough "mechanical margin" to withstand years of operation, expansions and day-to-day handling without developing widespread hidden faults.

On compliance, we aligned cable structures and architectures with IEC, ITU-T, TIA/EIA and ISO/IEC 11801, and worked with the customer's Australian design partners to meet local building and data centre rules on fire ratings, routing and penetrations. The system passed both technical and compliance reviews, which is the minimum requirement for a site that is expected to run for many years.

From this Sydney campus project, a few reasons to work with Hengtong stand out:

• Complete coverage from OSP to in-rack patching, with one responsible supplier
• Designs built from standard building blocks but customised to each campus
• Support that runs from early design through manufacturing, pre-termination, on-site guidance, testing and documentation
• Experience with international data centre projects, including Australia, so we share a common technical language with local consultants and operators
• Scalable production and quality systems to support multi-phase deployments with consistent performance

 

For operators, choosing a supplier is really choosing a long-term partner. This project shows the kind of role Hengtong can play in Australian data centre builds.