How Do You Choose the Right Fiber Count for Your Fiber Optic Cable Order?

2026-10-15T11:21:58+08:00

How Do You Choose the Right Fiber Count for Your Fiber Optic Cable Order? fiber count

A buyer's guide to choosing the right fiber count for fiber optic cable orders: counting day-one demand, spare capacity, construction limits, cost and common mistakes, with real 2–288 core outdoor and 2–48 core indoor ranges.

A buyer's question that arrives before the price does

Every fiber optic cable RFQ starts the same way. A buyer sends a route description, a distance, maybe a drawing, and then the manufacturer asks the first real question: how many fibers do you need in this cable? The answer, usually written as "fiber count" or "core count" on a datasheet, shapes everything that follows β€” the cable construction, the outer diameter, the drum weight, the price per kilometre, the pulling plan, the splicing schedule and how many years the same route can absorb network growth before anyone has to touch it again.

It sounds like a simple number. In practice it is the single most consequential decision on the order form, and it is also the one buyers most often get wrong in both directions. Order too few cores and the route has to be re-pulled in three years, at a cost dominated by civil works and labour rather than by the cable. Order too many and you pay for fibers that sit dark in a duct for a decade, plus a larger drum, a heavier pull, and a bigger splice closure at every joint.

This guide is written for the person who actually has to fill in that number on an RFQ: a telecom operator planning an access route, an ISP rolling out FTTH, a distributor stocking cable for resale, a system integrator wiring a campus, or a data center contractor sizing a backbone. We will work through how fiber count is defined, how to count what you actually need, how much spare capacity makes sense, how the number interacts with cable construction and mechanical limits, and where buyers most commonly miscalculate. Along the way we will use real product figures from Nexora's cable range β€” an outdoor G.652D loose-tube cable available in 2 to 288 cores and an indoor distribution cable available in 2 to 48 cores β€” so every recommendation is anchored to a configuration you can actually order.

What fiber count actually means β€” and what it does not

On a datasheet, fiber count is the number of individual optical fibers inside one cable sheath. A 24-core cable contains 24 separate glass fibers, each capable of carrying its own optical channel. That is the whole definition. But three misconceptions around this simple number cause most of the trouble.

First, fiber count is not the number of network connections. A single fiber is not a single link. With bidirectional transmission on one fiber, with wavelength-division techniques, or simply with the transmit and receive pairing of conventional transceivers, the relationship between "services delivered" and "fibers consumed" varies by architecture. A duplex link between two switches consumes two fibers β€” one to transmit, one to receive. A simplex bidirectional link consumes one. If your project plan says "48 connections at the far end", the fiber count you order may be 48, or 96, or something else entirely, depending on how those connections are terminated. Count fibers at the termination panel, not services on a network diagram.

Second, fiber count is not chosen once for the whole network. Different segments of the same project legitimately get different counts. A backbone route feeding several distribution points might justify 144 cores, while the spur leaving each distribution point to serve one street might need only 12. The distribution cable you break out at a floor ODF inside a building is typically a 2 to 48 core construction, while the outdoor cable entering the building may carry far more. Treating fiber count as one network-wide number is how projects end up both over-provisioned in one segment and starved in another.

Third, fiber count is not a statement about performance per fiber. A 12-core cable and a 288-core cable built on the same G.652D single mode fiber deliver the same attenuation per kilometre β€” typically ≀ 0.36 dB/km at 1310 nm and ≀ 0.22 dB/km at 1550 nm for this fiber class, with exact values confirmed on the specification sheet for your order. The count changes how many parallel paths you have; it does not change the quality of each path. That quality is set by the fiber type, which is a separate decision β€” if you are still choosing between fiber classes, see our comparison of G.657 versus G.652D fiber and our guide to single mode versus multimode fiber.

Start from the far end: count the fibers you actually need today

The reliable way to size a cable is to stop thinking about the cable and start thinking about the termination points. Fiber count is bought at the far end of the route: every fiber in the cable exists because something at a patch panel, an ODF, a splice closure or an equipment rack will eventually hold a connector or a splice for it. So the counting exercise begins at those panels, not at the cable drum.

Walk the route in your plan and list every termination point the cable will serve. For each point, write down how many fibers are actually landed there on day one. This is your illuminated fiber demand β€” the fibers that will carry traffic when the project goes live. For a straightforward access route with one headend and three distribution points, this might be as small as 12 fibers. For a building backbone serving eight floors with 48-port panels on each floor, it might be 384 β€” at which point you are not buying one cable, you are buying a backbone plus branches, and each gets its own count.

Three practical rules keep this count honest:

  • Count at the panel, not at the service. As discussed above, the mapping between services and fibers depends on your transmission architecture. The panel is where fibers physically land, so the panel is the unit of counting.
  • Count duplex applications as two fibers. Most conventional single mode links between transceivers use one fiber per direction. Until you have confirmed that your equipment plan uses bidirectional optics, every "link" on your diagram should be conservatively counted as two fibers.
  • Count the spares at the panel too. Many operators land spare fibers at the ODF even when the fibers are dark, so that a new service can be lit without opening a splice closure. If your maintenance practice lands spares at panels, those landed spares are part of your day-one fiber demand.

One worked example makes this concrete. Suppose you are ordering cable for an outdoor route from a headend to a street cabinet, and the cabinet serves 24 FTTH drop connections, each terminated as a conventional two-fiber link, plus one management link for the cabinet itself. Your day-one demand at the cabinet panel is 49 fibers. If your network design lands no spare fibers at the cabinet, and the next segment of the route continues onward to a second cabinet with 18 more connections, the cable passing through to that second cabinet must carry the first cabinet's demand plus the second's. This is why fiber count is planned route by route, not project-wide: the cable serving two cabinets in series carries both demands.

Once this day-one number is written down β€” say, 49 β€” the remaining question is how much headroom to add on top of it, and that is where the real judgement lies. The next section takes it on directly.

How much spare capacity is enough? A framework, not a formula

Spare capacity is where fiber count decisions get argued about, because the right answer depends on ownership horizon, growth expectations and the cost of touching the route again. There is no universal percentage, but there is a structured way to reason about it. The purpose of spare fibers is to make sure that the most expensive part of adding capacity later β€” accessing the route β€” never has to be repeated. Digging a trench, pulling cable through a congested duct, or arranging aerial access across somebody else's poles costs orders of magnitude more than the incremental price of a few extra cores ordered today.

Start by asking three questions about the route.

Question one: who owns the route's future? If you are a network operator that will operate this route for its full design life, you carry the cost of every future upgrade, so spare capacity is your own insurance and you can size it rationally. If you are a contractor delivering a turnkey project against a fixed scope, the customer owns the future, and over-provisioning is a cost you cannot recover β€” spec exactly what the contract says, plus only the spares the contract's own growth clauses require.

Question two: how accessible is the route? Spare capacity matters in proportion to how painful a re-pull would be. A direct-buried rural route crossing a river, a highway or private land is effectively frozen once installed; buyers commonly provision generous spare capacity on such routes for that reason. The same logic applies to routes in dense urban ducts that are already crowded. By contrast, a short indoor run inside your own building, where pulling a second distribution cable next year is a half-day job, justifies modest spares. This is one reason the outdoor cable family at Nexora extends to 288 cores while the indoor distribution family tops out at 48: the economics of spare capacity differ by orders of magnitude between those environments.

Question three: what does growth actually look like here? Look at your own network's history. If the services on similar routes doubled in five years, doubling your fiber demand in the calculation is not optimism β€” it is trend-following. If this is a first-time deployment with no comparable history, a common and defensible approach is to size the cable for the day-one demand plus a fixed growth allowance agreed with your planning team, and to lean on the splicing architecture (feeding several smaller cables from one larger one at a distribution point) rather than on a single enormous count.

As a reference frame drawn from real orders rather than theory: small access spurs are commonly ordered in the 2 to 12 core range, where demand is known and stable; campus and metro distribution segments typically sit between 24 and 48 cores; building and data center backbones frequently order 72 to 144 cores to cover multi-floor growth; and trunk routes feeding multiple distribution points are specified at 144 cores and above, up to the 288-core ceiling of the outdoor loose-tube family described here. These are typical practice bands, not rules β€” your own demand count and route economics should always override them.

One warning belongs in this section: spare fibers are only spare fibers if they survive the whole route intact. If you order 96 cores but splice through half of them at an intermediate closure, or land them at panels you do not control, your real spare capacity at the far end is much smaller than the datasheet number. Spare capacity is a property of the route plan, not just of the cable specification.

How fiber count changes the cable itself

A core count is not printed on an arbitrary cable. It drives the physical construction, and the construction drives the mechanical behaviour, the drum weight and the installation plan. Understanding this relationship helps you sanity-check a quotation that quotes a count you did not expect.

In a loose-tube outdoor construction β€” the family used for Nexora's outdoor G.652D single mode cable β€” fibers are grouped into gel-filled tubes, and the tubes are stranded around a central strength member. At low counts, a compact design with a single small tube or a few fibers per tube keeps the outer diameter tight. As the count rises, more tubes are added around the strength member, the core becomes physically larger, and the sheath follows. The published figures for the outdoor family illustrate the progression: cables in the 2 to 12 core range typically measure 9.0–10.5 mm in diameter and 85–110 kg/km, while 72 to 144 core builds run 12.0–14.5 mm and 150–210 kg/km, and the 216 to 288 core trunk constructions reach 14.5–16.5 mm and 210–260 kg/km. These are typical values for the standard construction with a steel wire strength member, tested to IEC 60794-1-21 tensile, crush and impact methods; the figures for your exact configuration are restated on the specification sheet issued with the order.

Mechanical limits move with the count as well. Short-term tensile ratings in the outdoor family typically progress from around 1,500 N at low counts to about 2,700 N at 216–288 cores, with long-term ratings from roughly 600 N to 1,000 N across the same span β€” typical figures for the steel-member build, with FRP constructions carrying roughly 20–25% lower tensile ratings at the same count. The practical consequence for your project is that a higher count is not just more glass: it is a heavier, stiffer cable that demands more attention to pulling tension, drum handling and bend-radius discipline during installation. A pulling plan written for a 12-core cable does not transfer to a 144-core pull without revision.

Inside buildings the picture is different. Nexora's indoor distribution cable carries 2 to 48 fibers in a tight-buffered or loose-tube construction under an LSZH jacket, sized for routing to floors, risers and data center rows. Here the count decision interacts with termination strategy more than with pulling mechanics. Tight-buffered fibers can be broken out and terminated directly at patch panels and ODFs without a splice enclosure, which is why distribution counts in buildings usually follow the panel plan quite closely β€” each fiber is meant to be handled individually, so you order what you will terminate, plus modest growth. If you are weighing jacket and construction choices for occupied spaces, our guide to LSZH cable and when to specify it covers that decision separately.

Outdoor backbone and indoor distribution follow different count rules

Buyers sometimes import outdoor sizing habits into indoor orders, and vice versa. The two environments reward different instincts, and confusing them produces predictable mistakes.

Outdoors, the dominant cost of getting the count wrong is re-access. The cable is buried, ducted or strung across spans once, and every future visit is charged at civil-works rates. That pushes outdoor counts upward: generous spares, larger constructions, and trunk routes sized for decades. The outdoor G.652D family supports this with a 2 to 288 core range, gel-filled water-blocking tubes, a choice of steel wire or FRP strength members and a black PE jacket rated for direct burial, duct and aerial deployment across a typical operating range of βˆ’40 Β°C to +70 Β°C. For routes with special access constraints, the same sizing logic extends to specialist constructions β€” ADSS for aerial routes where the cable must carry itself on a span, or armored builds where rodent or mechanical risk dominates β€” and those constructions carry their own count ceilings.

Indoors, the dominant cost of getting the count wrong is termination and tray congestion, not re-access. Pulling a second 24-core distribution cable through a riser next year is routine; pulling one 96-core cable to avoid that, and then finding tray space and slack management for it, may be the harder problem. That is why indoor counts stay bounded β€” 2 to 48 cores in the distribution family, with the indoor tight-buffered family covering 2 to 144 cores where riser and plenum backbones genuinely need higher densities β€” and why the sizing advice indoors is "follow the panel plan, add modest growth, and branch where branching is cheap".

There is also a count-adjacent decision that differs by environment: the break-out point. Outdoors, a 144-core trunk typically continues as a trunk until it reaches a distribution point, where it is spliced to several smaller cables. Indoors, a 48-core distribution cable is often terminated fiber by fiber at a single ODF. The first architecture is a splicing plan; the second is a termination plan. Your fiber count should be chosen knowing which plan applies, because the architectures waste spare fibers differently β€” a splice-through consumes fibers permanently, while an unterminated landed fiber at an ODF remains available.

When more cores do not help: fiber count and the link budget

A subtle failure mode in fiber count planning is treating cores as a substitute for optical budget. They are unrelated quantities, and a project that runs short on one sometimes tries to fix the problem with the other.

The link budget of any single fiber path is set by attenuation, splice loss, connector loss and margin β€” not by how many neighboring fibers share the sheath. On a 40 km outdoor span built on G.652D fiber, fiber attenuation alone consumes roughly 14.4 dB at 1310 nm and about 8.8 dB at 1550 nm, based on the typical ≀ 0.36 dB/km and ≀ 0.22 dB/km figures for this fiber class; splice loss, connector loss, cable margin and temperature allowances are added on top, and the exact values for your order are confirmed on the specification sheet. If a path does not close its budget, ordering a cable with more cores does not close it β€” the fix lies in fiber class, transmission equipment, route length or an intermediate splice point. See our G.657 versus G.652D comparison for the fiber-class side of that decision.

The converse is also worth stating: extra cores do no harm to the budget of the fibers you actually light. A dark fiber adjacent to a lit one does not degrade it. So over-provisioning costs money and handling, but it never costs optical performance. When a planner hesitates between two counts, the risk profile is asymmetric in a useful way β€” under-provisioning risks a civil-works-scale retrofit, while over-provisioning risks capital tied up in dark glass. That asymmetry is the honest reason spare capacity skews positive on hard-to-access routes.

What actually drives price as the count rises

Buyers often expect the price of a cable to scale linearly with fiber count. Quotations rarely behave that way, and understanding why helps you negotiate and budget more accurately.

Three cost components move with count. The fiber itself is the most obvious: more cores means more fibre kilometre, and at low counts this dominates the increment. The construction cost is less obvious: moving from a compact low-count core to a multi-tube stranded core changes the stranding process, the sheath diameter and the material consumption per kilometre, so the cable grows more expensive per core as the build changes. The handling cost arrives at the ends of the journey: bigger drums, higher drum weights (from roughly 85–110 kg/km at low counts up to 210–260 kg/km at 216–288 cores in the outdoor family, as typical values), heavier freight, and longer termination and splicing time at each end.

The practical implication: the marginal cost of the next few cores is steepest when it pushes the order across a construction boundary β€” for example from a compact 12-core build to a 24-core multi-tube build β€” and flattest within a build. If your day-one demand plus growth allowance lands near such a boundary anyway, the case for rounding up to the larger build is strongest right there, because you capture the extra capacity while paying mostly for fiber rather than for a new construction. Conversely, if growth would take you far past the boundary, it may be cheaper to plan a second cable on the route later than to buy one very large cable now.

MOQ behaves the same way. In the outdoor family, standard core counts typically start from 10 km per order, while custom core counts or non-standard sheath options are quoted from 20 km β€” typical reference figures, confirmed per configuration with the quotation. In the indoor distribution family, common builds start from around 1,000 m per configuration, with trial quantities available for a first order. If your route is short but your count is unusual, the MOQ of a custom build β€” not the unit price β€” may decide whether standardizing on a common count is the cheaper path. Our guide to MOQ and lead time for fiber optic orders covers that negotiation in detail.

The mistakes we see most often β€” and how to avoid them

After processing fiber count questions across hundreds of RFQs, the same handful of errors keeps appearing. None of them are exotic; all of them are avoidable at the desk, before the order is placed.

  • Counting services instead of fibers. The network diagram shows 24 links; the panel needs 48 fibers. Always convert to physical fibers at the termination point before writing the count on the RFQ.
  • Ignoring the middle of the route. A cable passing through an intermediate closure carries the demand of everything beyond that point. Route-by-route counting, not segment-by-segment counting, is what keeps series-fed routes sized correctly.
  • Zero spares on frozen routes. The cheapest fibre in a project is the core you ordered but did not need immediately β€” provided the route was the expensive-to-reopen kind. Direct-buried routes and congested ducts deserve deliberate, documented spare capacity.
  • Spares that never survive to the far end. As noted earlier, splice-throughs and intermediate landings consume nominal spares. Plan spare capacity against the route architecture, and state which fibers must remain through-spliced.
  • Copying a count from a previous project. Networks, contracts and growth histories differ. A count that was generous on one route can be a retrofit on the next.
  • Choosing a custom count when a standard one fits. Custom counts carry longer scheduling and higher MOQs. If your demand rounds to a standard construction without losing meaningful capacity, standardizing usually shortens delivery β€” standard configurations in the outdoor family are typically dispatched within 15–20 working days after order confirmation, with custom counts scheduled individually.
  • Forgetting the installation consequences of a high count. A 288-core trunk is a different physical object from a 12-core spur: heavier, stiffer, harder to pull. Make sure the pulling plan, drum handling and bend-radius limits in your method statement match the count you ordered.

One further note on documentation: whatever count you settle on, put the reasoning in the RFQ itself β€” day-one demand, growth allowance, route architecture and the through-splice plan. Quotations come back faster and more accurately when the supplier can see how the number was derived. If you are assembling a full RFQ, our checklist of what information to include in a fiber optic cable RFQ lists the rest of the fields.

A worked sizing walk-through you can copy

To close the technical half of this guide, here is the sequence in compact form β€” the same sequence our team applies when a buyer asks us to check a count before ordering.

  • Step 1 β€” Termination inventory. List every panel, ODF and closure the cable serves. For each, write the number of fibers landed on day one. Convert all links to physical fibers (duplex links count as two).
  • Step 2 β€” Route architecture. Mark which segments are series-fed and which are branched. The count of each cable equals the sum of day-one demand it carries beyond each point, not the demand of any single destination.
  • Step 3 β€” Growth allowance. Apply a documented growth figure per segment β€” from your network's own history where one exists, or from an agreed planning assumption where it does not. Note it explicitly so the count can be defended later.
  • Step 4 β€” Access penalty check. Classify the route: frozen (direct burial, congested duct, third-party poles), difficult (long aerial spans, road crossings) or open (own building, accessible trays). Let the classification push spare capacity up or down, and record why.
  • Step 5 β€” Construction fit. Match the resulting count to a real construction. Outdoors: 2–288 cores in the loose-tube G.652D family. Indoors: 2–48 cores in the distribution family, or the tight-buffered riser/plenum family up to 144 cores for genuine backbone demand. Round up only where the rounding crosses a construction boundary you would accept anyway.
  • Step 6 β€” Feasibility pass. Check mechanical limits against the installation method β€” tensile ratings against pulling loads, bend radii (typically 20 Γ— outer diameter in installation and 10 Γ— in service for the outdoor family) against tray and duct geometry, operating temperature against the route environment. These are typical values; confirm the figures for your exact configuration on the specification sheet before finalizing the method statement.
  • Step 7 β€” Confirmation. Send the count with its derivation to the manufacturer. A supplier who receives a reasoned count can confirm the construction, the strength member (steel wire where tensile demand is high, FRP where a non-metallic build is required), the sheath and the drum plan in one pass.

How we confirm a fiber count with you before production

Sizing guidance is only useful if the supplier behind it actually checks the number before committing a production slot. At Nexora, fiber count confirmation is a standard step between RFQ and order, and it works like this.

When your RFQ arrives with a core count and route description, our team first matches the count against the constructions we actually build β€” the 2 to 288 core range of the outdoor G.652D loose-tube family and the 2 to 48 core range of the indoor distribution family, plus the tight-buffered indoor family for higher riser densities. If the count fits a standard build, we confirm the sheath, strength member and drum lengths against your deployment method; if it does not, we come back with the nearest standard count that covers your demand and the schedule difference of a custom build, so the decision is made with both options visible.

Second, we flag the interactions the count creates: the mechanical values that move with count β€” the tensile progression from roughly 1,500 N at low counts to about 2,700 N at 216–288 cores in the steel-member outdoor build, with FRP builds typically 20–25% lower, all typical figures tested to IEC 60794-1-21 methods β€” and the bend-radius and temperature limits your installation plan must respect. The figures restated on your order's specification sheet are the governing ones; the values in this article describe typical ranges by core count, subject to specification sheet confirmation.

Third, we confirm the order-level practicalities that count affects: drum lengths (2 km, 4 km and 6 km standards in the outdoor family, with custom drum lengths available for OEM and ODM orders), jacket printing, packing, and the delivery schedule β€” typically 15–20 working days for standard outdoor configurations after order confirmation, and 7 to 15 working days for standard indoor distribution builds, with custom counts scheduled individually. Every Nexora fiber optic cable carries a 12-month warranty from shipment date covering manufacturing defects, with claims handled against the reel number, test report and fault description.

The background that makes this checking worthwhile: Nexora Communications has manufactured fiber optic cable and connectivity products in Shenzhen since 2014, running a 16,000+ mΒ² facility with 12 production lines, 30+ QC stations covering incoming material through packaging inspection, an annual capacity of 2.8M+ units, 3,500+ product configurations and more than 500 completed OEM/ODM projects delivered to customers in 50+ countries. Fiber count questions sit at the centre of that daily work β€” they are the first field on nearly every cable order we process, which is why this guide describes the same sequence we apply in practice rather than a theoretical framework.

Frequently asked questions

Does a higher fiber count change the attenuation of each fiber?

No. Attenuation is a property of the fiber and the construction quality, not of how many fibers share the sheath. All cores in the outdoor G.652D family are specified at the same fiber class β€” typically ≀ 0.36 dB/km at 1310 nm and ≀ 0.22 dB/km at 1550 nm, with exact per-order values on the specification sheet. What changes with count is the cable's diameter, weight and mechanical ratings.

Can I mix single mode and multimode fibers in one cable?

Mixed-count hybrid constructions exist, but they complicate splicing, testing and inventory, and most buyers avoid them. The cleaner pattern is one cable per fiber class: for example, a G.652D single mode outdoor cable for the backbone and a separate OM3/OM4 multimode distribution cable inside the data center, as covered in our single mode versus multimode guide.

How do I know whether my count needs a custom build?

Check the demand figure from your termination inventory against standard ranges. If it falls inside 2–288 cores outdoors or 2–48 cores indoors, a standard construction covers it β€” counts within those ranges do not require custom tooling, only configuration. Counts above the standard ceiling, unusual fiber-type mixes, or non-standard sheath requirements are the cases where a custom build and its longer schedule come into play.

Is it cheaper to order one 96-core cable now or two 48-core cables as demand grows?

It depends almost entirely on route access, not on the cable price. On a frozen route β€” direct burial, congested duct, third-party poles β€” one larger cable ordered now is usually cheaper in total cost of ownership, because a second pull later means civil works. In your own building or an accessible tray, two smaller pulls timed to real demand often win, and they avoid paying to handle a heavy high-count cable you did not yet need.

What information should I send with my fiber count when I request a quote?

Send the count, the route length, the deployment method (direct burial, duct or aerial), the termination plan at each end, the strength member preference if your route dictates one (steel wire for higher tensile demand, FRP for non-metallic builds), and the drum lengths you can handle at site. With those six facts, a manufacturer can confirm the construction, mechanical limits, packing and schedule in a single pass β€” the same checklist our RFQ information guide expands in full.

Do spare fibers need any maintenance while they are dark?

Dark fibers need the same physical protection as lit ones β€” the sheath, closures and panels must stay sealed and within bend-radius limits β€” but they consume no optical budget and require no transmission equipment. The one maintenance task that matters is documentation: keep the through-splice plan and panel records current so that when a spare is finally lit, its path is known without re-testing the whole route.

What does the warranty cover on a high-count cable?

The same terms apply across the count range: Nexora fiber optic cables carry a 12-month warranty from shipment date, covering manufacturing defects such as out-of-specification attenuation or construction faults on the delivered reels, with claims raised against the reel number and test report. It does not cover installation damage or third-party civil work. For a broader view of what warranty terms do and do not protect, see our 12-month warranty explainer.

Get your fiber count confirmed before you order

If you have a route plan and a working count, send it over before you commit the order. Tell us the core count, fiber type, route length and deployment method, and we will confirm the matching construction β€” outdoor G.652D loose-tube in 2 to 288 cores, indoor distribution in 2 to 48 cores, or the tight-buffered indoor family for riser and plenum backbones β€” along with the mechanical limits, drum plan and delivery schedule for your configuration. Every cable ships with reel-level test documentation and a 12-month warranty from shipment date, and sample requests are welcome when you want to verify the construction before a bulk order.