DAC vs Fibre - When Copper Stops Being the Answer

Quick answer

This one is decided by distance, not preference. Under about seven metres a passive direct-attach copper cable wins on every axis - cost, simplicity and heat. A DAC link starts at $8.54 against roughly $19.98 for two short-range optics before you buy any fibre. Past copper's ceiling you need optics and fibre, or an active optical cable - which reaches far further while still arriving as one factory-matched object.

Prices updated · live figures update every 4-5 hours

The copper path - passive direct-attach cables

Passive direct-attach copper, by length. A DAC is a fixed-length cable with the transceivers moulded on - you buy the whole run at once rather than a cable plus two modules.

LengthBrandTypePrice
0.3m10GtekSFP+ passive$9.99Buy
0.5mipolexSFP+ passive$8.54Buy
0.5mCable MattersSFP+ passive$9.22Buy
0.5mipolexSFP+ passive$9.49Buy
0.5mH!FiberSFP+ passive$9.69Buy
0.5m10GtekSFP+ passive$9.99Buy
0.5m10GtekSFP+ passive$9.99Buy
0.5mH!FiberSFP+ passive$9.99Buy
0.5mCable MattersSFP+ passive$13.99Buy
1mH!FiberSFP+ passive$9.99Buy
1mipolexSFP+ passive$11.04Buy
1m10GtekSFP+ passive$12.79Buy
1mH!FiberSFP+ passive$12.99Buy
1mipolexSFP+ passive$12.99Buy
1m10GtekSFP+ passive$14.99Buy
1mCable MattersSFP+ passive$14.99Buy
1.2m10GtekSFP+ passive$15.49Buy
1.5m10Gtek4 seller compat claimsSFP+ passive$15.49Buy
2mH!FiberSFP+ passive$11.99Buy
2m10GtekSFP+ passive$13.59Buy
2mipolexSFP+ passive$13.99Buy
2mH!FiberSFP+ passive$14.99Buy
2m10GtekSFP+ passive$16.99Buy
3mH!FiberSFP+ passive$15.29Buy
3mH!FiberSFP+ passive$16.99Buy
3mCable MattersSFP+ passive$18.99Buy
3m10GtekSFP+ passive$18.99Buy
5mCable MattersSFP+ passive$16.79Buy
5mCable MattersSFP+ passive$23.99Buy
7mH!FiberSFP+ passive$23.39Buy
7mH!FiberSFP+ passive$25.99Buy
7mipolexSFP+ passive$27.99Buy
7mQSFPTEKSFP+ passive$34.99Buy

The middle path - active optical cables

Every "compatible with" phrase below is the seller's own wording, recorded as a claim and badged. We have not tested any module against any switch.

ModuleTypeBrandSeller compatibility claimsPrice
SFP+ AOC 1mAOC10Gteknone stated$13.29Buy
SFP+ AOC 1mAOC10Gteknone stated$18.99Buy
SFP+ AOC 10mAOC10Gteknone stated$19.99Buy
SFP+ AOC 5mAOC10Gteknone stated$21.99Buy
SFP+ AOC 10mAOC10Gteknone stated$23.99Buy
SFP+ AOC 20mAOC10Gteknone stated$27.99Buy

The fibre path - short-range multimode optics

Every "compatible with" phrase below is the seller's own wording, recorded as a claim and badged. We have not tested any module against any switch.

ModuleTypeBrandSeller compatibility claimsPrice
SFP+ 10GBASE-SR module10GBASE-SRQSFPTEKnone stated$9.99Buy
SFP+ 10GBASE-SR module10GBASE-SR10Gteknone stated$11.89Buy
SFP+ 10GBASE-SR module10GBASE-SR10Gteknone stated$16.99Buy

The fibre path - long-range single-mode optics

Every "compatible with" phrase below is the seller's own wording, recorded as a claim and badged. We have not tested any module against any switch.

ModuleTypeBrandSeller compatibility claimsPrice
SFP+ 10GBASE-LR module10GBASE-LR10Gteknone stated$12.59Buy
SFP+ 10GBASE-LR module10GBASE-LR10Gteknone stated$17.99Buy

The decision is almost entirely about distance

Unusually for a comparison page, this one has a simple rule and very few exceptions. Under about seven metres, use a direct-attach copper cable. Over it, you need optics or an active optical cable. Cost, complexity and heat all favour copper, and the only reason to move away from it is that copper physically stops working.

That seven-metre figure is the practical ceiling of passive direct-attach copper. It is a specification limit rather than a quality one: reputable makers do not sell longer passive cables because a longer passive cable would not meet the standard. If you find one advertised well beyond that, be suspicious rather than pleased.

So the honest framing of this comparison is not "which is better". It is "how far apart are the two ends", and everything else follows from that answer.

What each path actually costs

A direct-attach cable is one purchase. The transceiver ends are moulded onto the cable at the factory, so there is nothing else to buy and nothing to match. The tables above show the current range across every length we stock.

The fibre path is three purchases: a module for each end plus the fibre patch lead between them. Even before the fibre, two modules cost several times what a direct-attach cable of usable length costs, and that gap is the single biggest reason short runs should never use fibre.

An active optical cable sits between the two commercially and behaves like a direct-attach cable operationally - one sealed object, ends fixed at the factory, no module matching. It costs more than passive copper and less than a full fibre build, and it reaches far beyond copper's ceiling.

For a two-machine build in one room, the cost difference between the copper and fibre paths is large enough that it will usually dominate the whole build price. That is worth checking against the live figures above rather than taking on trust, because it is exactly the kind of ratio that shifts.

Complexity, which is the underrated axis

Copper's real advantage is not the price, it is that there is nothing to get wrong. One object, two ends, no decisions. Nothing to match, no fibre type to select, no polarity to think about, and effectively no compatibility question because there is no separate module for a switch to reject.

Fibre asks you to get several things right simultaneously. The module type must match at both ends - short-range with short-range, long-range with long-range. The fibre must match the module: multimode for short-range optics, single-mode for long-range. Mixing those is the usual reason a link fails to come up, and the failure gives you very little to go on.

Then there is the compatibility question, which applies to separate modules in full force. A switch enforcing vendor coding can refuse a module that does not identify as an approved part, and third-party modules are programmed to identify as a named vendor's part precisely to route around that. Those compatibility statements are seller claims, badged as such in our tables, and we have not tested any of them.

Fibre also demands more physical care. The end faces are precision optical surfaces, dust caps exist for a reason, and a fibre that has been dragged across a floor may not work and may not obviously look damaged. Twinax copper is far more forgiving of ordinary handling, though it is stiff and has a real bend radius at longer lengths.

Where fibre earns it

Between buildings, or across a large one. This is the case fibre exists for and copper cannot address at any price - long-range single-mode optics are measured in kilometres.

Where electrical isolation matters. Fibre carries no current, so it does not create a path between the electrical systems at either end. In a home this rarely matters; between separate structures it can matter a great deal.

Where the run is permanent and the endpoints are not. A structured fibre run can outlive several generations of the equipment plugged into it, because upgrading means changing the modules rather than the cable in the wall. A direct-attach cable is a fixed pairing of cable and ends, so replacing either means replacing all of it.

And where the cable has to be thin or the bend radius tight. Fibre patch leads are far more flexible than long twinax, which is a genuine practical consideration in dense cable management even at short distances.

Active optical cables, the middle option people forget

An AOC gives you most of fibre's reach with most of copper's simplicity, and it is routinely left out of this comparison.

Physically it is a fibre with the optics permanently attached at both ends - so it arrives as one object, the ends are matched at the factory, and there is no module for a switch to reject. Operationally you treat it exactly like a direct-attach cable that happens to reach much further.

The trade is the same one direct-attach copper makes: the length is fixed at purchase and nothing about the assembly can be changed later. If a module fails you replace the whole cable rather than one end.

For a run that is too long for passive copper but well inside one building, an AOC is very often the right answer and it is cheaper and simpler than assembling a fibre link from parts. Our current AOC prices are in the table above alongside the copper.

Where to go next

DAC cables by length · transceivers and the compatibility maze · the two-machine build uses the copper path · SFP+ versus 10GBASE-T is the decision upstream of this one.

Frequently asked questions

When should I use fibre instead of a DAC cable?

When the run is longer than passive copper can serve - around seven metres is the practical ceiling. Under that, a direct-attach cable is cheaper, simpler and has no module compatibility question. Fibre also wins where electrical isolation matters, where the cable run is permanent and the equipment is not, or where you need a thin, flexible cable.

How far can a DAC cable go?

Around seven metres for passive direct-attach copper, which is where the longest cables on the market sit. It is a specification limit rather than a quality one - a longer passive cable would not meet the standard, so reputable makers do not sell one. Past that you want an active optical cable or separate optics and fibre.

Is fibre faster than a DAC cable?

No. Both carry a 10GbE link at 10GbE. The choice is about distance, cost and complexity, not about speed. Anyone selling you fibre on performance grounds for a short run is selling on the wrong axis.

What is an AOC and where does it fit?

An active optical cable is a fibre with the optics permanently attached at both ends, so it behaves like a direct-attach cable that reaches much further - one object, factory-matched ends, no module to be rejected. For a run too long for copper but inside one building it is often the right answer, and simpler than assembling a fibre link from separate parts.

Do I need to match fibre types?

Yes, and it is the usual reason a fibre link fails to come up. Short-range optics need multimode fibre, long-range optics need single-mode, and the module type has to be the same at both ends. A direct-attach or active optical cable removes this entirely because both ends are matched at the factory.

Are DAC cables affected by vendor lock?

Much less than separate modules, but not literally never - a DAC's moulded ends still identify themselves to the port, so a switch enforcing vendor coding could in principle refuse one. It is reported far less often, and for a link between two machines whose cards you chose yourself the question effectively does not arise.

Sources

Specification claims on this page were checked against the pages below on 4 August 2026. Where a widely-repeated figure did not survive that check, we report what the source says now and say so in the text rather than repeating the familiar number.

  1. NVIDIA (Mellanox) - ConnectX-3 EN product line and lifecycle status (nvidia.com, 2026)
    Used for: that ConnectX-3 has reached end-of-life status with NVIDIA, and the MCX311A / MCX312A / MCX314A part-number family. EOL means no new firmware or driver work from the maker; it does not mean the card stops working, and the page says so in those terms.
  2. The Linux kernel project - mlx4, mlx5, ixgbe and i40e in-tree Ethernet drivers (kernel.org, 2026)
    Used for: that mlx4 (ConnectX-3), mlx5 (ConnectX-4 and later), ixgbe (Intel 82599/X520, X540, X550) and i40e (X710) ship in the mainline Linux kernel. This is why an end-of-life card can still be a sane purchase, and it is a documented fact rather than a forum claim.
  3. DatacenterDisk - Live catalogue - NIC, DAC and transceiver listings (datacenterdisk.com, 2026-08-11)
    Used for: every price on this page, read from our catalogue at request time. Seeded 11 August 2026: 24 network cards across 6 models, 24 DAC cables, 8 transceivers. Every card is split tier 1 (the listing is that part) or tier 2 (a third-party board asserting that silicon).
  4. DatacenterDisk - Sourcing note - power figures and compatibility claims (datacenterdisk.com, 2026-08-11)
    Used for: We publish no per-port wattage for either cabling path. The relationship - 10GBASE-T draws more power and runs hotter than SFP+ direct attach - is documented at product level and is stated here. The magnitude is not, because we have not fetched and logged a source we can cite for it. Separately: every 'compatible with' phrase shown on this site is stored as a seller claim at tier 2 and badged wherever it renders. We have not tested any module against any switch.
Share:WhatsAppFacebookPost on X