DP to HDMI Cable for 4K60: The Bandwidth Budget

October 31, 2026

By ZONG Editorial Team | ZONG technical guide

Quick answer: The build rule: DP 1.2 source with HBR2 at 21.6 Gbps, HDMI 2.0 sink at 18 Gbps, target 4K60 8-bit RGB. Because the conversion spends the top of the HDMI budget on overhead, the cable must be short, well shielded and correctly terminated. Keep passive assemblies to 2 m at 4K60 and move to an active part beyond that length. If the room only needs 1080p or 4K30, a passive part stays the cheapest correct answer.

Start with the budget, not the box. A 4K60 frame at 8-bit RGB needs about 17.8 Gbps of HDMI 2.0 payload once blanking is counted, and the link offers 18 Gbps. That leaves roughly one per cent of margin before any cable loss is added. A passive DP to HDMI cable has to live inside that margin, and most assemblies above 2 m cannot.

The arithmetic, done once

DisplayPort 1.2 supplies 21.6 Gbps over four lanes, so the source has more raw capacity than the sink will take. HDMI 2.0 accepts 18 Gbps over three TMDS lanes plus a clock channel. The bridge chip has spare input bandwidth and no spare output bandwidth, which inverts the instinct that a cable only needs to match the source.

That asymmetry is why a cable rated for the DisplayPort side can still fail. The DP link trains and reports a clean result in DPCD, while the HDMI link behind it runs with no error margin.

The 0.2 Gbps of headroom

4K60 at 8-bit RGB uses a 594 MHz pixel clock, and with standard blanking the payload lands near 17.8 Gbps against an 18 Gbps ceiling. Deep colour at 10-bit needs about 22.3 Gbps, which HDMI 2.0 cannot carry at 4K60 in RGB.

Where DisplayPort 1.4 changes nothing

A DP 1.4 source at HBR3 offers 32.4 Gbps, but an HDMI 2.0 sink still caps at 18 Gbps. Buying a faster source does not buy a longer passive cable. Only a full HDMI 2.1 chain at 48 Gbps changes that, and it needs both ends.

What a 4K60 cable must do electrically

At 6 Gbps per lane the assembly is an RF problem rather than a wire problem. Three parameters decide the result: insertion loss at 3 GHz, intra-pair skew, and shield bonding at each connector shell.

Insertion loss and the eye

A 2 m passive assembly typically loses 8 dB to 10 dB at 3 GHz, a 3 m assembly 12 dB to 15 dB. The sink equaliser has a finite budget, so once the eye closes the link fails training or steps down a timing.

Termination and the ground path

A shield bonded at one end only turns the cable into an antenna. The symptom is not always a dead link; more often it is an intermittent drop when a monitor arm moves, which is hard to reproduce on a bench.

Build options compared

Reach and shield construction matter more than the printed specification.

Why 28 AWG is not automatically worse

Gauge sets DC loss, but at 3 GHz the dielectric and the connector dominate. A well made 28 AWG assembly with a solid dielectric can beat a 24 AWG assembly whose foam dielectric crushes under a cable tie.

Fibre hybrid where it is justified

A fibre hybrid carries the high speed pairs optically and copper only for power and the sideband channel. It is the practical answer past 10 m at 4K60, and it is directional, so the source end must be labelled at the factory.

Build	4K60 reach	Shield	Best use
28 AWG passive	2 m	Foil plus braid	Desk and dock
24 AWG passive	2.5 m	Braid, both ends	Short rack runs
Active redriver	5 m	Braid, both ends	Wall and ceiling
Fibre hybrid	20 m	Full braid	Control room
Adapter plus lead	Depends on lead	Varies	Mixed stock

Real installations

Two builds where the budget, not the catalogue, decided the part.

Rack to bench, 2.4 m

A test bench ran 4K60 from a DP 1.2 card to an HDMI 2.0 capture device. A 24 AWG passive assembly at 2.4 m held the link with zero errors for 48 hours. The same assembly at 3.2 m produced 4K30 within four minutes.

Training room, 65 inch panel

The panel sat 3.8 m from the lectern and a passive run failed on warm-up. An active 5 m cable held 4K60 through a 41 degrees C afternoon, and the installer gained a service loop at each end.

Frequently asked questions

Is 4K60 possible on a passive DP to HDMI cable?

At 2 m, usually. At 3 m, only with a well built assembly and a tolerant sink. Beyond that the 0.2 Gbps of headroom cannot absorb the loss and the link steps down.

Why does the picture look fine but the refresh rate reads 30 Hz?

The link trained at a lower rate. The desktop still renders, so nothing looks broken, but the source has chosen a timing the cable can carry. Check the refresh rate in display settings.

Does HDR change the cable requirement?

It can. 4K60 at 10-bit needs about 22.3 Gbps, above the HDMI 2.0 ceiling. HDR at 4K60 on that link means 4:2:0 chroma, which changes the picture but not the cable job.

Are active cables directional?

Yes. The retimer sits at the source end and the shell is marked. Reversing an active cable gives no picture at all rather than a degraded one, which makes the mistake quick to diagnose.

Should I buy a certified cable?

Certification confirms the assembly met a test limit when it was tested. That is useful evidence, but for 4K60 at 2 m to 3 m ask for the insertion loss figure and the gauge.

Next step

Send the source, the sink and the measured route length, and we will return the build that keeps 4K60 inside budget, including the gauge, the shield construction and whether an active part is needed. Ask at https://zh-zong.com/contact/, email zongcarrie@zh-zong.com or call +86 13543026126.