---
title: "Why is there a chip beside every Thunderbolt port?"
topic: "Retimers and redrivers"
answer: "Because 20 Gb/s doesn’t survive the trip across the board. Simulated on ordinary laptop board, one USB4 lane passes the test at the port for 70 mm of trace with no chip. A 14-inch MacBook Pro’s ports sit about 130 mm from its processor. A redriver stretches the reach to 210 mm, and a retimer passed at every length tried, out to 410 mm."
url: "https://nearfieldnotes.com/notes/retimers/"
site: "Near Field Notes"
author: "Michael Walker"
note: 42
family: "Signal integrity"
published: "2026-11-16"
revised: "2026-11-16"
revision: 1
video: "https://nearfieldnotes.com/videos/retimers/"
---

Note 42 · [Signal integrity](https://nearfieldnotes.com/families/si/) · Rev 1 · [Watch the video · 2:44](https://nearfieldnotes.com/videos/retimers/)

Retimers and redrivers

# Why is there a chip beside every Thunderbolt port?

**Short answer:** Because 20 Gb/s doesn’t survive the trip across the board. Simulated on ordinary laptop board, one USB4 lane passes the test at the port for 70 mm of trace with no chip. A 14-inch MacBook Pro’s ports sit about 130 mm from its processor. A redriver stretches the reach to 210 mm, and a retimer passed at every length tried, out to 410 mm.

A redriver is an analog boost: it puts back the high frequencies the board took. A retimer reads every bit and sends it again from a fresh transmitter.

USB4 allows a redriver only with a retimer between it and the connector, so a redriver can’t be the last chip before a USB4 connector.

**Note 42’s opening figure: The eye at the port.** No chip, as the trace from the SoC grows

![Note 42’s opening figure: The eye at the port](https://nearfieldnotes.com/notes/retimers/fig-opening.png)

*What the figure shows:* A simulated 20 gigabit per second USB4 signal at a laptop's USB-C port, with no chip on the way, drawn as eye diagrams of 8,000 bits laid over each other, two bits wide, for three lengths of board trace from the processor. A bar in the middle of each eye marks the 240 millivolt opening USB4's test at the port asks for. At 10 mm the eye is 363 millivolts open and passes. At 70 mm it is 245 and still passes. At 130 mm it is 164 millivolts, and the bar runs into the traces: it fails.

Simulated: 8,000 bits of one 20 Gb/s lane at the laptop’s USB-C port, laid over each other, two bits wide. The bar is the 240 mV of opening USB4’s test asks for. The number is the eye’s height over a million bits, the way the test counts it. The chart ends with all three side by side.

## One USB4 lane, from the SoC to the port

*20 Gb/s, 50 ps a bit*

A Thunderbolt 4 or USB4 port moves 40 Gb/s over two lanes each way, 20 Gb/s on each. One bit lasts 50 ps.

This page follows one lane from a laptop’s SoC along a board trace to its USB-C port, then down a 0.8 m cable to a receiver. Every part on the way loses more at high frequencies than at low ones. The trace loses the most, and more the longer it gets: 1.4 dB per inch at 10 GHz on ordinary laptop board.

When there is a chip, it sits 12 mm of trace before the connector. The trace is swept from 10 to 410 mm, three times: with no chip, with a redriver and with a retimer.

**Fig. 42a: The lane, and its two tests.**

![Fig. 42a: The lane, and its two tests](https://nearfieldnotes.com/notes/retimers/fig-42a.png)

*What the figure shows:* Figure 42a: the simulated lane, drawn as a diagram. Top row, inside the laptop: the SoC, a board trace of 10 to 410 mm, the spot for a chip, 12 mm of trace, and the USB-C port. Test 1 is taken at the port. A cable with 7.5 dB of loss runs from the port back along the bottom row to a receiver, and Test 2 is taken there, after the receiver's equalizer.

Drawn for this page. Test 1 is at the laptop’s own connector, Test 2 at the far end of the cable.

**The numbers**

| Part | Loss at 10 GHz |  |
|---|---|---|
| SoC package and coupling capacitors | 1.5 dB | assumed |
| Board trace | 1.4 dB per inch | 1.4 on ordinary laptop board, 2.0 on plain FR-4 |
| Two vias and the ESD part | 1.1 dB | 0.8 and 0.3 |
| USB-C connector | 1.0 dB |  |
| Cable, 0.8 m | 7.5 dB | the limit for a passive USB4 Gen 3 cable |
| The chip's package | 1.0 dB | retimer only |

USB4 tests a laptop’s port twice. At its own connector, with no equalizer, the eye must open 240 mV, 120 each side of zero, over a million bits, and stay 27 ps wide at one error in 10¹² bits. At the far end of a cable with 7.5 dB of loss, after the spec’s reference equalizer, it must open 98 mV and stay 20 ps wide.

An [eye diagram](https://nearfieldnotes.com/glossary/#g-eye-diagram) lays the bits over each other. The test counts far more bits than a picture of a few thousand can show, so the opening is drawn here as two lenses: white where no bit crosses in 10¹² bits, pale where one may cross by 10⁶.

The cross in the middle is the test. Its upright bar has to fit inside the pale lens, and its crossbar inside the white one.

**Fig. 42b: Reading the test on an eye.**

![Fig. 42b: Reading the test on an eye](https://nearfieldnotes.com/notes/retimers/fig-42b.png)

*What the figure shows:* Figure 42b: the eye at the port after 70 mm of board with no chip, simulated, twice. Left: 8,000 bits laid over each other, two bits wide, with two lens-shaped outlines in the opening, the larger where a bit may cross by a million bits and the smaller where none crosses by 10 to the 12th. Right: the middle bit enlarged and drawn as those two openings, white inside the smaller lens and pale inside the larger, with a cross on it: an upright bar 240 millivolts tall and a crossbar 27 picoseconds wide. The bar fits the pale lens, 245 millivolts tall, and the crossbar fits the white one, 30.8 picoseconds wide, so the port passes.

Simulated, at the port after 70 mm of board with no chip. Left: 8,000 bits, with the two openings outlined. Right: the middle bit enlarged. Every eye from here on is drawn the second way, in its lane’s color.

## With no chip, the eye closes by 75 mm

*245 mV at 70 mm*

With nothing between the SoC and the port, the eye at the port is 363 mV after 10 mm of trace. It shrinks as the trace grows, and clears 240 mV for the last time at 70 mm, with 245 mV. At 75 mm it is 239.6 mV.

70 mm is a loss of 7.5 dB at 10 GHz from the SoC to the connector. That is the figure USB4 recommends for a laptop, 7.5 dB, in a section marked informative. The model wasn’t tuned to land there.

At the far end the same lane lasts a little longer, to 105 mm, so the port’s own test fails first. At 130 mm the eye at the port is 164 mV, and the far end fails too.

**Fig. 42c: The eye at the port, no chip.**

![Fig. 42c: The eye at the port, no chip](https://nearfieldnotes.com/notes/retimers/fig-42c.png)

*What the figure shows:* Figure 42c: the eye height at the port with no chip, against the length of board trace from the SoC, simulated. It falls from 363 millivolts at 10 mm, crosses the 240 millivolts USB4 asks for between 70 and 75 mm, and is 164 at 130 mm and 68 at 210 mm.

Simulated. Height over a million bits at the USB-C connector, with no equalizer.

**The numbers**

Height over a million bits, width at one error in 10¹². The port test asks for 240 mV and 27 ps.

| Trace | Height | Width | Loss to the port at 10 GHz | Port test |
|---|---|---|---|---|
| 10 mm | 362.6 mV | 34.6 ps | 4.2 dB | Passes |
| 30 mm | 322.4 mV | 33.7 ps | 5.2 dB | Passes |
| 50 mm | 277.8 mV | 32.3 ps | 6.4 dB | Passes |
| 60 mm | 258.7 mV | 32.1 ps | 6.9 dB | Passes |
| 65 mm | 248.5 mV | 31.9 ps | 7.2 dB | Passes |
| 70 mm | 245.3 mV | 30.8 ps | 7.5 dB | Passes |
| 75 mm | 239.6 mV | 30.7 ps | 7.7 dB | Fails |
| 80 mm | 230.1 mV | 30.5 ps | 8.0 dB | Fails |
| 100 mm | 193.6 mV | 29.9 ps | 9.1 dB | Fails |
| 130 mm | 163.9 mV | 28.5 ps | 10.8 dB | Fails |
| 160 mm | 123.9 mV | 27.4 ps | 12.4 dB | Fails |
| 210 mm | 68.0 mV | 24.0 ps | 15.2 dB | Fails |

## A redriver puts back the high frequencies

*8 dB at 10 GHz*

A redriver is an analog chip whose gain rises with frequency. It boosts the highs the board lost more than the lows, so each pulse comes out closer to the shape it left with.

At 130 mm the board from the SoC to the port loses 10.8 dB at 10 GHz, half the bit rate. The best setting there is 8 dB of boost, and together they lose 2.8 dB. The far receiver’s own equalizer handles the rest.

With it, the 130 mm lane passes both tests: 365 mV at the port, and 156 mV and 25.1 ps at the far end.

**Fig. 42d: Gain against frequency, 130 mm.**

![Fig. 42d: Gain against frequency, 130 mm](https://nearfieldnotes.com/notes/retimers/fig-42d.png)

*What the figure shows:* Figure 42d: gain against frequency for the 130 mm lane, simulated. The board from the SoC to the port loses more at higher frequencies: 10.8 dB at 10 GHz, half the bit rate, and 18.5 dB at 20. The redriver's setting adds 8 dB at 10 GHz and little at low frequencies. Together they lose 2.8 dB at 10 GHz.

Simulated, with the redriver at the setting the sweep picked for this length. The dotted line is 10 GHz, half of 20 Gb/s.

**The numbers**

The 130 mm lane, with the redriver at the 8 dB setting the sweep picked there.

| Frequency | Board, SoC to port | Redriver | Both |
|---|---|---|---|
| 0 GHz | 0.0 dB | 1.4 dB | 1.4 dB |
| 10 GHz | −10.8 dB | 8.0 dB | −2.8 dB |
| 20 GHz | −18.5 dB | 10.3 dB | −8.2 dB |

## The redriver runs out at 210 mm

*30.3 ps to 20.2 ps*

Each extra length needs more boost: 14 dB by 210 mm. Each time, the far-end eye comes out a little narrower, from 30.3 ps at 10 mm to 20.2 ps at 210 mm. At 215 mm it is 19.9 ps, under the 20 ps the far-end test asks for.

Past 210 mm the best setting passes at a few lengths, 220, 225, 245 mm, and fails at the rest. Its reach here is the last length before the first failure.

**Fig. 42e: The far-end eye’s width.**

![Fig. 42e: The far-end eye’s width](https://nearfieldnotes.com/notes/retimers/fig-42e.png)

*What the figure shows:* Figure 42e: the width of the eye at the far end of the cable, after the reference equalizer, against the length of board trace, simulated. With a redriver it narrows from 30.3 picoseconds at 10 mm to 20.2 at 210 mm, then wavers around the 20 picoseconds USB4 asks for and falls to 14.8 at 410 mm. With a retimer it stays at 30.3 picoseconds at every length.

Simulated, at the far end of the cable after the reference equalizer. Width at one error in 10¹² bits.

**The numbers**

At the far end, after the reference equalizer. Width at one error in 10¹², height over a million bits. Past 210 mm the redriver's best setting passes at a few lengths and fails at the rest; its reach is the last length before the first failure.

| Trace | Redriver: width | height | setting | Retimer: width | Redriver, both tests |
|---|---|---|---|---|---|
| 10 mm | 30.25 ps | 223 mV | 3 dB | 30.31 ps | Passes |
| 50 mm | 28.82 ps | 206 mV | 5 dB | 30.32 ps | Passes |
| 100 mm | 26.50 ps | 173 mV | 7 dB | 30.32 ps | Passes |
| 130 mm | 25.14 ps | 156 mV | 8 dB | 30.32 ps | Passes |
| 150 mm | 23.91 ps | 147 mV | 9 dB | 30.32 ps | Passes |
| 200 mm | 20.71 ps | 187 mV | 14 dB | 30.32 ps | Passes |
| 205 mm | 20.44 ps | 180 mV | 14 dB | 30.32 ps | Passes |
| 210 mm | 20.18 ps | 175 mV | 14 dB | 30.32 ps | Passes |
| 215 mm | 19.90 ps | 168 mV | 14 dB | 30.32 ps | Fails |
| 220 mm | 20.46 ps | 194 mV | 15 dB | 30.32 ps | Passes |
| 225 mm | 20.18 ps | 185 mV | 15 dB | 30.32 ps | Passes |
| 230 mm | 19.91 ps | 178 mV | 15 dB | 30.32 ps | Fails |
| 245 mm | 20.31 ps | 188 mV | 16 dB | 30.32 ps | Passes |
| 260 mm | 19.27 ps | 166 mV | 16 dB | 30.32 ps | Fails |
| 300 mm | 18.77 ps | 161 mV | 18 dB | 30.32 ps | Fails |
| 350 mm | 17.79 ps | 133 mV | 19 dB | 30.32 ps | Fails |
| 410 mm | 14.82 ps | 90 mV | 19 dB | 30.32 ps | Fails |

What narrows it? The model can switch parts off one at a time. At 230 mm, take away the crosstalk, the receiver’s noise or the redriver’s own noise, and the width moves by 0.2 ps at most. Noise isn’t what stops it.

Take away the [jitter](https://nearfieldnotes.com/glossary/#g-jitter) and the eye opens from 19.9 ps to 28.4 ps. But jitter costs the retimer just as much: 30.3 ps with it, 40.1 ps without.

Take all of them away and the redriver’s eye is 29.2 ps wide, against the retimer’s 40.9 ps. What’s left is smear: each pulse spread into its neighbors, more than the boost and the far receiver’s equalizer can take back. The smear eats the room the jitter needs.

**Fig. 42f: What closes the redriver’s eye, at 230 mm.**

![Fig. 42f: What closes the redriver’s eye, at 230 mm](https://nearfieldnotes.com/notes/retimers/fig-42f.png)

*What the figure shows:* Figure 42f: the width of the far-end eye at 230 mm of board, simulated, with parts of the model switched off one at a time. Redriver: 19.9 picoseconds as simulated; 20.0 with no crosstalk; 20.0 with no receiver noise; 20.1 with no noise from the redriver; 28.4 with no jitter; 29.2 with none of these. Retimer: 30.3 as simulated and 40.9 with none of these. USB4 asks for 20.

Simulated, far-end width at one error in 10¹² bits, with parts of the model switched off. The dotted line is the 20 ps USB4 asks for.

**The numbers**

At the far end, after the reference equalizer. Width at one error in 10¹², height over a million bits. The redriver keeps the 15 dB setting the sweep picked at this length.

| At 230 mm | Redriver: width | height | Retimer: width | height |
|---|---|---|---|---|
| As simulated | 19.91 ps | 177.8 mV | 30.32 ps | 175.4 mV |
| No crosstalk | 20.03 ps | 178.2 mV | 30.50 ps | 176.0 mV |
| No receiver noise | 19.95 ps | 177.9 mV | 30.41 ps | 175.7 mV |
| No redriver noise | 20.07 ps | 178.4 mV | 30.32 ps | 175.4 mV |
| No jitter | 28.42 ps | 223.4 mV | 40.08 ps | 206.3 mV |
| None of these | 29.23 ps | 226.3 mV | 40.92 ps | 209.6 mV |

## A retimer starts the signal over

*175 mV and 30.3 ps at every length*

A retimer is a receiver and a transmitter in one package. It equalizes what arrives, finds the clock in it, decides whether each bit is a 1 or a 0, and sends the bits again at full swing.

So the far end only ever sees the 12 mm of trace and the connector after the retimer. The board before it is the retimer’s problem: its own receiver has to read every bit right. USB4 trains each hop on its own, the SoC to the retimer and the retimer to the far end.

**Fig. 42g: Inside each chip.**

![Fig. 42g: Inside each chip](https://nearfieldnotes.com/notes/retimers/fig-42g.png)

*What the figure shows:* Figure 42g: what each chip does, as block diagrams. A redriver is one block: boost the high frequencies. A retimer is four in a row: equalize, find the clock, decide each bit, and send it again.

Drawn for this page. A redriver never decides a bit. A retimer decides every one.

The far-end eye comes out the same at every length: 175 mV tall and 30.3 ps wide. At the port it is 388 mV.

What changes is the eye inside the retimer, after its own equalizer. At 410 mm, the end of the sweep, it is down to 26.5 mV and 19.3 ps at one error in 10¹², and still open. Kandou says its KB8001 retimer can sit about 40 cm from the SoC on low-cost board.

A bit-by-bit run of 8,000 bits found no wrong bits at the retimer at any length. A real retimer that gets a bit wrong sends the wrong bit on, cleanly.

**Fig. 42h: The retimer’s eyes, inside and at the far end.**

![Fig. 42h: The retimer’s eyes, inside and at the far end](https://nearfieldnotes.com/notes/retimers/fig-42h.png)

*What the figure shows:* Figure 42h: the retimer's eyes at three lengths of board, simulated, drawn as their openings. Top row, inside the retimer after its own equalizer, the opening at one error in 10 to the 12th: 10 mm, 492 millivolts tall and 36.0 picoseconds wide; 210 mm, 129 millivolts tall and 29.7 picoseconds wide; 410 mm, 27 millivolts tall and 19.3 picoseconds wide. Bottom row, at the far end of the cable, with USB4's test drawn as a cross: 10 mm, 177 millivolts and 30.3 picoseconds; 210 mm, 175 millivolts and 30.3 picoseconds; 410 mm, 175 millivolts and 30.3 picoseconds, the same eye every time.

Simulated. Inside the retimer the numbers are the white lens’s height and width: a bit it gets wrong goes out wrong. At the far end they are the test’s.

**The numbers**

Inside the retimer, height and width at one error in 10¹². At the far end, height over a million bits and width at 10¹².

| Trace | Inside: height | width | equalizer | Far end: height | width |
|---|---|---|---|---|---|
| 10 mm | 491.5 mV | 36.0 ps | 0 dB | 176.8 mV | 30.3 ps |
| 50 mm | 381.6 mV | 35.4 ps | 1 dB | 175.3 mV | 30.3 ps |
| 100 mm | 276.1 mV | 34.0 ps | 2 dB | 175.3 mV | 30.3 ps |
| 130 mm | 223.2 mV | 33.0 ps | 3 dB | 175.3 mV | 30.3 ps |
| 200 mm | 136.2 mV | 30.2 ps | 5 dB | 175.3 mV | 30.3 ps |
| 250 mm | 100.3 mV | 27.3 ps | 6 dB | 175.3 mV | 30.3 ps |
| 300 mm | 67.5 mV | 25.3 ps | 7 dB | 175.3 mV | 30.3 ps |
| 350 mm | 47.2 mV | 23.0 ps | 9 dB | 175.3 mV | 30.3 ps |
| 400 mm | 29.6 mV | 20.9 ps | 9 dB | 175.3 mV | 30.3 ps |
| 410 mm | 26.5 mV | 19.3 ps | 11 dB | 175.3 mV | 30.3 ps |

## How far each one reaches

*70, 210 and 410+ mm*

Side by side, the three lanes pass both tests to 70 mm, 210 mm and 410 mm, the end of the sweep.

A 14-inch MacBook Pro’s three ports sit about 11 to 14 cm from the middle of its M1 Pro, in a straight line, scaled from iFixit’s photo. A routed trace is longer. At 130 mm the lane with no chip fails both tests, and the redriver and the retimer both pass.

The board matters as much as the chip. On plain FR-4, at 2.0 dB per inch, the reaches drop to 45 mm and 145 mm. The retimer still passes at 410 mm, but its own eye is 19.6 ps wide at 300 mm and only 1.8 ps at 410.

**Fig. 42i: Where each lane passes both tests.**

![Fig. 42i: Where each lane passes both tests](https://nearfieldnotes.com/notes/retimers/fig-42i.png)

*What the figure shows:* Figure 42i: the lengths of board trace at which each lane passes both of USB4's tests, simulated in 5 mm steps from 10 to 410 mm. On ordinary laptop board: no chip to 70 mm, a redriver to 210 mm and again at a few lengths past it, a retimer at every length to 410 mm, the end of the sweep. A line marks 130 mm. On plain FR-4: no chip to 45 mm, a redriver to 145 mm, a retimer at every length to 410.

Simulated in 5 mm steps. A filled step passes both tests at that length. The retimer also needs its own input eye open. The line is 130 mm.

## Move the port yourself

*10 to 410 mm*

Slide the port away from the SoC. Each column is one lane and each row one of USB4’s tests. A lane passes a test when the cross fits its eye.

**Note 42’s explorer: Six eyes.** 10.8 dB lost on the way to the port at 10 GHz

![Note 42’s explorer: Six eyes](https://nearfieldnotes.com/notes/retimers/fig-explore.png)

No chip

Redriver

Retimer

At the port: 240 mV and 27 ps

*What the figure shows:* No chip, at the port: the eye's opening, with USB4's test drawn as a cross

164 mV, 28.5 ps Fails

*What the figure shows:* Redriver, at the port: the eye's opening, with USB4's test drawn as a cross

365 mV, 30.3 ps Passes

*What the figure shows:* Retimer, at the port: the eye's opening, with USB4's test drawn as a cross

388 mV, 35.1 ps Passes

At the far end: 98 mV and 20 ps

*What the figure shows:* No chip, at the far end: the eye's opening, with USB4's test drawn as a cross

84 mV, 25.4 ps Fails

*What the figure shows:* Redriver, at the far end: the eye's opening, with USB4's test drawn as a cross

156 mV, 25.1 ps Passes

*What the figure shows:* Retimer, at the far end: the eye's opening, with USB4's test drawn as a cross

175 mV, 30.3 ps Passes

Each length is its own run of the model, in 10 mm steps, with the settings the sweep picked for it. Height over a million bits, width at one error in 10¹². Past 210 mm the redriver still passes at a few lengths, 220 mm among them. The retimer’s own input eye isn’t shown here; Fig. 42h has it.

## USB4 rules out a redriver at the port

*Up to two retimers*

USB4 settles which chip goes nearest the port. A redriver may sit inside the laptop, but only with a retimer after it: “If a Router Assembly contains Linear Re-drivers, a Re-timer shall be placed between the Linear Re-driver and the USB Type-C connector.” A port may have up to two retimers.

No chip at all is allowed too. USB4 judges the port at its connector, and its 7.5 dB is a recommendation. A port close enough to its SoC needs nothing.

A redriver working alone belongs to slower links. USB 3.2’s appendix on repeaters covers redrivers on the board with no retimer, for 10 Gb/s and a host allowance of 8.5 dB at 5 GHz. The iPhone 15 Pro’s port runs at 10 Gb/s, and iFixit lists no such chip beside it. Framework kept Thunderbolt off its 12-inch laptop; its CEO wrote that “Thunderbolt requires much more expensive retimers.”

**Fig. 42j: What may sit before a USB4 port.**

![Fig. 42j: What may sit before a USB4 port](https://nearfieldnotes.com/notes/retimers/fig-42j.png)

*What the figure shows:* Figure 42j: what may sit between the SoC and a USB4 port. Nothing: allowed if the port passes its test. A retimer: allowed. A redriver and then a retimer: allowed. A redriver alone, last before the port: not allowed in USB4.

From the USB4 specification, version 2.0: the Electrical Layer’s rule on redrivers and retimers, and section 2.1.1.6 on how many retimers a port may have.

## Where the chips are

*Laptops, a Mac mini, a phone, a cable*

**Fig. 42k: Two retimers on a laptop board.**

![Fig. 42k: Two retimers on a laptop board](https://nearfieldnotes.com/notes/retimers/fig-42k.png)

*What the figure shows:* Figure 42k: a photograph of part of a laptop's circuit board, dark blue, scanned flat. Two square black chips marked intel JHL8040R, Intel's Thunderbolt 4 retimers, are outlined, each 6 mm on a side and surrounded by small capacitors. Beside each one sits a silver metal can. Photograph by Rjluna2, released as CC0 on Wikimedia Commons.

Two Intel JHL8040R, outlined. Photograph by Rjluna2 on Wikimedia Commons, released as CC0. The source doesn’t name the laptop. Scale from the chip’s 6 mm package.

Retimer

### Intel JHL8040R

**Field marks.** A square chip about 6 mm on a side, about 1 to 2 cm inboard of a USB-C port, one per port, ringed by small capacitors. Intel’s are marked JHL8040R, and Parade’s PS8830 is the same 6 mm.

**Look-alike.** The port’s power delivery controller, often close by: on the M1 MacBooks a TI CD3217. Teardown lists name it apart from the retimer.

**Range.** USB4 and Thunderbolt ports on laptops and tablets, and the plugs of long Thunderbolt cables.

**Habitat.** Common since 2019, when Intel moved the Thunderbolt controller into its Ice Lake processors and the JHL8040R took its place by the ports. Intel’s recommended price is $2.90.

An M1 Mac mini’s logic board (2020)

*What the figure shows:* Photograph by Sonic8400, CC BY-SA 4.0: part of an M1 Mac mini's logic board, from 2020. On the right, the M1: a silver package with the Apple logo, its two memory chips beside it. On the left, outlined, two square black chips marked intel JHL8040R, one above the other, each with a small crystal just above it: the Mac mini's two Thunderbolt retimers.

The M1 at right, and at left its two Intel JHL8040R retimers, outlined, each with a small crystal just above it. The retimers are 6 mm square. Photo: [Sonic8400, via Wikimedia Commons](https://commons.wikimedia.org/wiki/File:M1_A13_comparison_MacMini9_1_M1.jpg), cropped, with outlines added. License: [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/), outlines included. Unlike this page’s figures, it isn’t under CC BY 4.0.

**Figure: Fast ports and the chips beside them.**

Fig. 42l

| Device | Fast ports | Beside them | Source |
|---|---|---|---|
| MacBook Air and 13-inch MacBook Pro, M1 (2020) | 2 Thunderbolt | Intel JHL8040R retimer, two | iFixit |
| 14-inch MacBook Pro, M1 Pro (2021) | 3 Thunderbolt 4 | Intel JHL8040R retimer, one inboard of each port | iFixit |
| MacBook Air, M2 (2022) | 2 Thunderbolt | An Apple-marked chip where the Intel retimers were. iFixit: “likely” a redriver; Tom’s Hardware: a retimer | iFixit, Tom’s Hardware |
| Surface Laptop 7, 13.8-inch (2024) | 2 USB4 | Parade PS8830 retimer, one inboard of each port | iFixit |
| Framework Laptop 13, AMD Ryzen 7040 (2023) | 2 USB4, 2 USB 3.2 | Kandou retimers on the two USB4 slots, an Analogix retimer on one USB 3.2 slot | Framework |
| Framework Laptop 12 (2025) | 4 USB 3.2 | No Thunderbolt: it “requires much more expensive retimers” | Framework’s CEO |
| iPhone 15 Pro (2023) | 1 USB 3, 10 Gb/s | None listed | iFixit |
| Apple Thunderbolt 4 Pro cable | In the plugs | 1.8 m: an Intel JHL7040 retimer. 1 m: none, a passive cable | ChargerLAB |

From teardowns and the makers’ own words, each named in Sources. A chip missing from a teardown’s list isn’t proof it isn’t there. Nobody has published what is inside the M2 MacBook Air’s Apple chip.

## What goes beside a fast port?

*Two questions*

It comes down to the speed of the link and the loss between the SoC and the connector.

Key to a fast USB-C port

### What sits beside it?

1. 1a 10 Gb/s or less, USB 3.2 Nothing if the port passes; otherwise a redriver can work alone
2. 1b 20 Gb/s a lane, USB4 or Thunderbolt Go to 2
3. 2a The port passes with nothing in the way No chip
4. 2b It doesn’t A retimer, last before the connector; a redriver only before it

USB4 recommends 7.5 dB or less at 10 GHz from the SoC to the connector, about 70 mm of ordinary laptop board in Fig. 42c. The test at the connector decides.

#### Conditions

- **Simulated:** Every eye, height, width and reach on this page comes from a statistical model of one lane, in the style of IEEE 802.3’s Annex 93A. None was measured on a bench.
- **Lane:** One USB4 Gen 3 lane, 20 Gb/s, 50 ps a bit. Transmitter 800 mV peak to peak with a 20 ps edge, and the best of USB4’s 16 presets at each length.
- **Board:** 1.4 dB per inch at 10 GHz, scaled from Intel’s measurements of a reference board at 8 GHz, 1.06 to 1.38 dB per inch. It is the weakest number here. Plain FR-4: 2.0 dB per inch, from TI’s rule of 0.2 dB per inch per GHz.
- **Fixed parts:** At 10 GHz: SoC package and coupling capacitors 1.5 dB (assumed), two vias 0.8, the ESD part 0.3, the connector 1.0.
- **Cable:** 0.8 m, 7.5 dB at 10 GHz, USB4’s figure for a Gen 3 passive cable.
- **Redriver:** Linear, 12 mm of trace before the connector: 0 to 19 dB of boost at 10 GHz in 1 dB steps, the best setting at each length, 3 mV rms of its own noise, linear to 1.2 V peak to peak.
- **Retimer:** 12 mm before the connector, behind 1.0 dB of its own package. Its receiver has the reference equalizer’s shape with 0 to 24 dB, and one feedback tap of up to 50 mV. Its transmitter is like the SoC’s.
- **Receiver:** At the far end, USB4’s reference equalizer: 0 to 9 dB and one decision-feedback tap of up to 50 mV. Crosstalk at −44 dB at 10 GHz, receiver noise, 0.01 UI rms of random jitter and 0.085 UI of bounded jitter.
- **Tests:** USB4 Table 3-7 at the port, no equalizer: 240 mV over 10⁶ bits and total jitter 0.46 UI at 10⁻¹², so 27 ps of eye. Table 3-8 at the far end: 98 mV over 10⁶ bits and total jitter 0.60 UI, so 20 ps.
- **Sweep:** 10 to 410 mm in 5 mm steps. A lane’s reach is the last length before its first failure. The retimer also needs its own input eye open.
- **Check:** A bit-by-bit run of 8,000 bits at every length, with the same settings, draws the eye pictures. The retimer decided every bit right. The pictures open a little wider than the numbers, which count rarer bits.

A model of one lane. Nothing on this page has been measured on a bench.

#### Limits

- The board’s loss is scaled, not measured. A lower-loss laminate or package moves every reach, and plain FR-4 cuts them to 45 mm and 145 mm.
- Loss only: no reflections from via stubs, the connector or the package, and crosstalk as one fixed level.
- The redriver is an ideal linear one, with one zero and two poles. Real redrivers’ curves differ.
- A model in the style of Annex 93A, not USB4’s compliance software.
- Past 210 mm the redriver passes at 220, 225, 245 mm and fails at the rest. The reach counts the first failure, at 215 mm.
- On FR-4 at 410 mm the retimer passes, but its input eye is 1.8 ps wide. That pass rests on the eye being open at all.
- The 130 mm is a straight-line distance scaled from a photo, good to about 30%. A routed trace is longer.
- Teardown lists aren’t complete. A chip missing from one isn’t proof it isn’t there.
- Power, delay and cost aren’t modeled.

#### Files

- sim/link.py The lane: its parts, the statistical eye, both tests, the sweep, the parts switched off one at a time, the bit-by-bit eyes and the numbers for this page

#### Sources

- [USB-IF, USB4 Specification, Version 2.0: Tables 3-7 and 3-8 (the tests at the port and the far end), section 3.1.1.1 (the recommended 7.5 dB, informative), the Electrical Layer’s rule on redrivers and retimers, and section 2.1.1.6 (up to two retimers a port)](https://www.usb.org/document-library/usb4r-specification-v20)
- [USB-IF, USB 3.2 Specification, Revision 1.1, Appendix E: redrivers and retimers on the board, and the 8.5 dB host allowance at 5 GHz](https://www.usb.org/document-library/usb-32-revision-11-june-2022)
- [USB-IF, USB Type-C Cable and Connector Specification, Release 2.4: a Gen 3 passive cable’s 7.5 dB at 10 GHz](https://www.usb.org/document-library/usb-type-cr-cable-and-connector-specification-release-24)
- [Intel, Electrical Characterization Design and Methodologies Guide, 337658-001, Tables 5-1 and 5-2: measured trace loss at 4 and 8 GHz](https://www.intel.com/content/dam/www/public/us/en/documents/guides/electrical-character-design-meth-guide-337658-rev001.pdf)
- [TI, TUSB1002A datasheet: trace loss of 0.2 dB per inch per GHz](https://www.ti.com/lit/ds/symlink/tusb1002a.pdf)
- [Intel, JHL8040R Thunderbolt 4 Retimer specifications: 6 × 6 mm package, launched Q3 2019, $2.90 recommended price](https://www.intel.com/content/www/us/en/products/sku/186251/intel-jhl8040r-thunderbolt-4-retimer/specifications.html)
- [SlashGear, 2019-05-28: Ice Lake’s built-in Thunderbolt 3 still needs a retimer at the ports](https://www.slashgear.com/intel-10th-gen-ice-lake-cpus-official-project-athena-sets-pc-blueprint-28578200)
- [Kandou, KB8001 announcement, 2021-06-02: a retimer about 40 cm from the SoC on low-cost board](https://www.eejournal.com/industry_news/kandou-unveils-first-device-in-usb-c-multiprotocol-retimer-product-family-for-usb4/)
- [Parade, PS8830 USB4 retimer: 6 × 6 mm package](https://www.paradetech.com/products/ps8830/)
- [iFixit, M1 MacBook teardowns: two Intel JHL8040R retimers, and the TI CD3217 power delivery controller](https://www.ifixit.com/News/46884/m1-mac)
- [iFixit, MacBook Pro 14-inch 2021 Chip ID, step 2: the JHL8040R retimers, and the photo the 11 to 14 cm is scaled from](https://www.ifixit.com/Teardown/MacBook+Pro+14-Inch+2021+Chip+ID/145718#s296569)
- [iFixit, MacBook Air (M2 2022) Logic Board and Chip Identification: the Apple chip, “likely” a redriver](https://www.ifixit.com/Teardown/Macbook+Air+(M2+2022)+Logic+Board+and+Chip+Identification/151816#s316005)
- [Tom’s Hardware: the same Apple chip, called a retimer](https://www.tomshardware.com/news/apple-amd-removes-intel-silicon)
- [iFixit, Surface Laptop 7 (13.8-inch) Chip ID, step 2: two Parade PS8830 retimers](https://www.ifixit.com/Teardown/Microsoft+Surface+Laptop+7+(13.8-inch)+Chip+ID/173921#s370876)
- [Framework, expansion card functionality on the Framework Laptop 13 (AMD Ryzen 7040): Kandou and Analogix retimers](https://knowledgebase.frame.work/en_us/expansion-card-functionality-on-framework-laptop-13-amd-ryzen-7040-series-SkrVx7gAh)
- [Framework community forum, 2025-04-10, Framework’s CEO on the Laptop 12 and retimer cost](https://community.frame.work/t/no-thunderbolt/67331/8)
- [iFixit, iPhone 15 teardown: the Pro’s USB 3 port at 10 Gb/s](https://www.ifixit.com/News/82867/iphone-15-teardown-reveals-software-lockdown)
- [iFixit, iPhone 15 Pro Max Chip ID: no retimer or redriver listed](https://www.ifixit.com/Teardown/iPhone+15+Pro+Max+Chip+ID/165320)
- [ChargerLAB, Apple Thunderbolt 4 Pro cable 1.8 m teardown: an Intel JHL7040 retimer](https://www.chargerlab.com/the-most-expensive-cable-teardown-of-apple-thunderbolt-4-pro-cable-1-8-m/)
- [ChargerLAB, Apple Thunderbolt 4 Pro cable 1 m teardown: passive, no retimer](https://www.chargerlab.com/teardown-of-apple-thunderbolt-4-pro-1m-cable-a2804/)
- [Wikimedia Commons, “Intel jhl8040r.jpg” by Rjluna2, CC0: the photograph in Fig. 42k and the note’s cover](https://commons.wikimedia.org/wiki/File:Intel_jhl8040r.jpg)
- [Wikimedia Commons, “M1 A13 comparison MacMini9 1 M1.jpg” by Sonic8400, CC BY-SA 4.0: the M1 Mac mini photograph and its two JHL8040R retimers](https://commons.wikimedia.org/wiki/File:M1_A13_comparison_MacMini9_1_M1.jpg)

What should the next note be? [Ask for one](https://nearfieldnotes.com/#ask-for-a-note)

---

Source: https://nearfieldnotes.com/notes/retimers/

Cite as: Michael Walker, “Retimers and redrivers”, Near Field Notes, Note 42, Rev 1, 16 Nov 2026. https://nearfieldnotes.com/notes/retimers/

Figures: free to reuse under CC BY 4.0 (https://nearfieldnotes.com/license/). Each one's image is linked under its name. Credit Michael Walker, Near Field Notes, and link the figure: this address plus #fig- and its number, like https://nearfieldnotes.com/notes/retimers/#fig-42a.
